Introduction

Imagine just as you’re sipping your Monday morning coffee and looking forward to a hopefully quiet week in the office, your Outlook dings and you see that your bank’s primary federal regulator is demanding the full input – regulatory report lineage for dozens of numbers on both sides of the balance sheet and the income statement for your latest financial report filed with the regulator. The full first day letter responses are due next Monday, and as your headache starts you remember that the spreadsheet owner is on leave; the ETL developer is debugging a separate pipeline; and your overworked and understaffed reporting team has three different ad hoc diagrams that neither match nor reconcile.

If you can relate to that scenario, or your back starts to tighten in empathy, you’re not alone. Artificial Intelligence (“AI”) driven data lineage for banks is no longer a nice-to-have. We at Perficient working with our clients in banking, insurance, credit unions, and asset managers find that it’s the practical answer to audit pressure, model risk (remember Lehman Brothers and Bear Stearns), and the brittle manual processes that create blind spots. This blog post explains what AI-driven lineage actually delivers, why it matters for banks today, and a phased roadmap Chief Data Officers (“CDOs”) can use to get from pilot to production.

Why AI-driven data lineage for banks matters today

Regulatory pressure and real-world consequences

Regulators and supervisors emphasize demonstrable lineage, timely reconciliation, and governance evidence. In practice, financial services firms must show not just who touched data, but what data enrichment and/or transformations happened, why decisions used specific fields, and how controls were applied—especially under BCBS 239 guidance and evolving supervisory expectations.

In addition, as a former Risk Manager, the author knows that he would have wanted and has spoken to a plethora of financial services executives who want to know that the decisions they’re making on liquidity funding, investments, recording P&L, and hedging trades are based on the correct numbers. This is especially challenging at global firms that operate in in a transaction heavy environment with constantly changing political, interest rate, foreign exchange and credit risk environment.

Operational risks that keep CDOs up at night

Manual lineage—spreadsheets, tribal knowledge, and siloed code—creates slow audits, delayed incident response, and fragile model governance. AI-driven lineage automates discovery and keeps lineage living and queryable, turning reactive fire drills into documented, repeatable processes that will greatly shorten the time QA tickets are closed and reduce compensation costs for misdirected funds. It also provides a scalable foundation for governed data practices without sacrificing traceability.

What AI-driven lineage and controls actually do (written by and for non-tech staff)

At its core, AI-driven data lineage combines automated scanning of code, SQL, ETL jobs, APIs, and metadata with semantic analysis that links technical fields to business concepts. Instead of a static map, executives using AI-driven data lineage get a living graph that shows data provenance at the field level: where a value originated, which transformations touched it, and which reports, models, or downstream services consume it.

AI adds value by surfacing hidden links. Natural language processing reads table descriptions, SQL comments, and even README files (yes they do still exist out there) to suggest business-term mappings that close the business-IT gap. That semantic layer is what turns a technical lineage graph into audit-ready evidence that regulators or auditors can understand.

How AI fixes the pain points keeping CDOs up at night

Faster audits: As a consultant at Perficient, I have seen AI-driven lineage that after implementation allowed executives to answer traceability questions in hours rather than weeks. Automated evidence packages—exportable lineage views and transformation logs—provide auditors with a reproducible trail.
Root-cause and incident response: When a report or model spikes, impact analysis highlights which datasets and pipelines are involved, highlighting responsibility and accountability, speeding remediation and alleviating downstream impact.
Model safety and feature provenance: Lineage that includes training datasets and feature transformations enables validation of model inputs, reproducibility of training data, and enforcement of data controls—supporting explainability and governance requirements. That allows your P&L to be more R&S. (a slogan used by a client that used R&S P&L to mean rock solid profit and loss.)

Tooling, architecture, and vendor considerations

When evaluating vendors, demand field-level lineage, semantic parsing (NLP across SQL, code, and docs), auditable diagram exports, and policy enforcement hooks that integrate with data protection tools. Deployment choices matter in regulated banking environments; hybrid architectures that keep sensitive metadata on-prem while leveraging cloud analytics often strike a pragmatic balance.

A practical, phased roadmap for CDOs

Phase 0 — Align leadership and define success: Engage CRO, COO, and Head of Model Risk. Define 3–5 KPIs (e.g., lineage coverage, evidence time, mean time to root cause) and what “good” will look like. This is often done during a evidence gathering phase by Perficient with clients who are just starting their Artificial Intelligence journey.
Phase 1 — Inventory and quick wins: Target a high-risk area such as regulatory reporting, a few production models, or a critical data domain. Validate inventory manually to establish baseline credibility.
Phase 2 — Pilot AI lineage and controls: Run automated discovery, measure accuracy and false positives, and quantify time savings. Expect iterations as the model improves with curated mappings.
Phase 1 and 2 are usually done by Perficient with clients as a Proof-of-Concept phase to show that the key feeds into and out of existing technology platforms can be done.
Phase 3 — Operationalize and scale: Integrate lineage into release workflows, assign lineage stewards, set SLAs, and connect with ticketing and monitoring systems to embed lineage into day-to-day operations.
Phase 4 — Measure, refine, expand: Track KPIs, adjust models and rules, and broaden scope to additional reports, pipelines, and models as confidence grows.

Risks, human oversight, and governance guardrails

AI reduces toil but does not remove accountability. Executives, auditors and regulators either do or should require deterministic evidence and human-reviewed lineage. Treat AI outputs as recommendations subject to curator approval. This will avoid what many financial services executives are dealing with what is now known as AI Hallucinations.

Guardrails include the establishment of exception processing workflows for disputed outputs and toll gates to ensure security and privacy are baked into design—DSPM, masking, and appropriate IAM controls should be integral, not afterthoughts.

Conclusion and next steps

AI data lineage for banks is a pragmatic control that directly addresses regulatory expectations, speeds audits, and reduces model and reporting risk. Start small, prove value with a focused pilot, and embed lineage into standard data stewardship processes. If you’re a CDO looking to move quickly with minimal risk, contact Perficient to run a tailored assessment and pilot design that maps directly to your audit and governance priorities. We’ll help translate proof into firm-wide control and confidence.

Introduction:

Custom code in Talend offers a powerful way to enhance batch processing efficiently by allowing developers to implement specialized logic that is not available through Talend’s standard components. This can involve data transformations, custom code as per use case and integration with flat files as per specific project needs. By leveraging custom code, users can optimize performance, improve data quality, and streamline complex batch workflows within their Talend jobs.

Talend Components:

Key components for batch processing as mention below:

• tDBConnection: Establish and manage database connections within a job & allow configuration with single connection to reuse within Talend job.
• tFileInputDelimited: For reading data from flat files.
• tFileRowCount: Reads file row by row to calculate the number of rows.
• tLoop: Executes a task automatically, based on a loop size.
• tHashInput, tHashOutput: For high-speed data transfer and processing within a job. tHashOutput writes data to cache memory, while tHashInput reads from that cached data.
• tFilterRow: For filtering rows from a dataset based on specified.
• tMap: Data transformation allows you to map input data with output data and enables you to perform data filtering, complex data manipulation, typecasting, and multiple input source joins.
• tJavaRow: It can be used as an intermediate component, and we are able to access the input flow and transform the data using custom Java code.
• tJava: It has no input or output data flow & can be used independently to Integrate custom Java code.
• tPreJob, tPostJob: PreJob start the execution before the job & PostJob at the end of the job.
• tDBOutput: Supports wide range of databases & used to write data to various databases.
• tDBCommit:It retains and verifies the alterations applied to a connected database throughout a Talend job, guaranteeing that it permanently records the data changes.
• tDBClose:  It explicitly close a database connection that was opened by a tDBConnection component.
• tLogCatcher: It is used in error handling within Talend job for adding runtime logging information. It catches all the exceptions and warnings raised by tWarn and tDie components during Talend job execution.
• tLogRow: It is employed in error handling to display data or keep track of processed data in the run console.
• tDie: We can stop the job execution explicitly if it fails. In addition, we can create a customized warning message and exit code.

Workflow with example:

To process the bulk of data in Talend, we can implement batch processing to efficiently process flat file data within a minimal execution time. We can read the flat file data & after the execution, we can process it to insert it into MySQL database table as a target & we can achieve this without batch processing. But this data flow will take quite a longer time to execute. If we use batch processing using the custom code, it takes minimal execution time to write the entire source file data into batch of records into MySQL database table at the target location.

Talend Job Design 

Solution:

• Establish the database connection at the start of the execution so that we can reuse.
• Read the number of rows in the source flat file using tFileRowCount component.
• To determine the batch size, subtract the header count from the total row count and then divide the number by the total batch size. Take the whole number nearby which indicates the total number of batch or chunk.
  

  Calculate the batch size from total row count

• Now use tFileInputDelimited component to read the source file content. In the tMap component, utilize the sequence Talend function to generate row numbers for your data mapping and transformation tasks. Then, load all of the data into the tHashOutput component, which stores the data into a cache.
• Iterate the loop based on the calculated whole number using tLoop
• Retrieve all the data from tHashInput component.
• Filter the dataset retrieved from tHashInput component based on the rowNo column in the schema using tFilterRow

Filter the dataset using tFilterRow

• If First Iteration is in progress & batch size is 100 then rowNo range will be as 1 to 100.
  If Third Iteration is in progress & batch size is 100 then rowNo range will be as 201 to 300.
  For example, if the value of current iteration is 3 then [(3-1=2)* 100]+1 = 201 & [3*100=300]. So final dataset range for the 3^rd iteration will be 201 to 300.
• Finally extract the dataset range between the rowNo column & write the batch data MySQL database table using tDBOutput
• The system uses the tLogCatcher component for error management by capturing runtime logging details, including warning or exception messages, and employs tLogRow to display the information in the execution console.
• Regarding performance tuning, tMap component that maps source data to output data, allows for complex data transformation, and offers unique join, first join, and all other join options for looking up data within the tMap component.
• The temporary data that the tHashInput & tHashOutput components store in cache memory enhances runtime performance.
• At the end of the job execution, we are committing the database modification & closing the connection to release the database resource.

Advantages of Batch Processing:

• Batch processing can efficiently handle large datasets.
• It takes minimal time to process the data even after data transformation.
• By grouping records from a large dataset and processing them as a single unit, it can be highly beneficial for improving performance.
• With the batch processing, it can easily scale to accommodate growing data volumes.
• It is particularly useful for operations like generating reports, performing data integration, and executing complex transformations on large datasets.

For more details: Get-started-talend-open-studio-data-integration

Note: Efficient Batch Processing in Talend-Part 1

Introduction:

Custom code in Talend offers a powerful way to enhance batch processing efficiently by allowing developers to implement specialized logic that is not available through Talend’s standard components. This can involve data transformations, custom code as per use case and integration with flat files as per specific project needs. By leveraging custom code, users can optimize performance, improve data quality, and streamline complex batch workflows within their Talend jobs.

Understand Batch Processing:

            Batch processing is a method of running high-volume, repetitive data within Talend jobs. The batch method allows users to process a bunch of data when computing resources are available, and with little or no user interaction.

Through batch processing, users gather and retain data, subsequently processing it during a designated period referred to as a “batch window.” This method enhances efficiency by establishing processing priorities and executing data tasks in a timeframe that is optimal.

Here, Talend job takes the total row count from source file then load the data from the flat file, processes it in a batch, provided input through context variable & then write the data into smaller flat files. This implementation made it possible to process enormous amounts of data more precisely and quickly than other implementation.

Batch processing is a method of executing a series of jobs sequentially without user interaction, typically used for handling large volumes of data efficiently. Talend, a prominent and extensively employed ETL (Extract, Transform, Load) tool, utilizes batch processing to facilitate the integration, transformation, and loading of data into data warehouse and various other target systems.

Talend Components:

Key components for batch processing as mention below:

• tFileInputDelimited, tFileOutputDelimited: For reading & writing data from/to files.
• tFileRowCount: Reads file row by row to calculate the number of rows.
• tLoop: Executes a task automatically, based on a loop size.
• tHashInput, tHashOutput: For high-speed data transfer and processing within a job. tHashOutput writes data to cache memory, while tHashInput reads from that cached data.
• tFilterRow: For filtering rows from a dataset based on specified.
• tMap: Use for data transformation which allow to map input data with output data along with use to perform data filtering, complex data manipulation, typecasting & multiple input source join.
• tJavaRow: It can be used as an intermediate component, and we are able to access the input flow and transform the data using custom Java code.
• tJava: It has no input or output data flow & can be used independently to Integrate custom Java code.
• tLogCatcher: It is used in error handling within Talend job for adding runtime logging information. It catches all the exceptions and warnings raised by tWarn and tDie components during Talend job execution.
• tLogRow: It is employed in error handling to display data or keep track of processed data in the run console.

Workflow with example:

To process the bulk of data in Talend, we can implement batch processing to efficiently process flat file data within a minimal execution time. We can read the flat file data & after execution, we can write it into a chunk of another flat file as a target & we can achieve this without batch processing. But this data flow will take quite a larger execution time to execute. If we use batch processing using the custom code, it takes minimal execution time to write the entire source file data into chunks of files at the target location.

Talend job design

Solution:

• Read the number of rows in the source flat file using tFileRowCount component.
• To determine the batch size, subtract the header count from the total row count and then divide the number by the total batch size. Take the whole number nearby which indicates the total number of batch or chunk.
  

  Calculate the batch size from total row count

• Now use tFileInputDelimited component to read the source file content. In the tMap component, utilize the sequence Talend function to generate row numbers for your data mapping and transformation tasks. Then, load all of the data into the tHashOutput component, which stores the data into a cache.
• Iterate the loop based on the calculated whole number using tLoop
• Retrieve all the data from tHashInput component.
• Filter the dataset retrieved from tHashInput component based on the rowNo column in the schema using tFilterRow
  

  Filter the dataset using tFilterRow

• If First Iteration is in progress & batch size is 100 then rowNo range will be as 1 to 100.
  If Third Iteration is in progress & batch size is 100 then rowNo range will be as 201 to 300.
  For example, if the value of current iteration is 3 then [(3-1=2)* 100]+1 = 201 & [3*100=300]. So final dataset range for the 3^rd iteration will be 201 to 300.
• Finally extract the dataset range between the rowNo column & write it into chunk of output target file using tFileOutputDelimited
• The system uses the tLogCatcher component for error management by capturing runtime logging details, including warning or exception messages, and employs tLogRow to display the information in the execution console.
• Regarding performance tuning, we have a tMap component that maps source data to output data, allows for complex data transformation, and offers unique join, first join, and all other join options for looking up data within the tMap component.
• The temporary data that the tHashInput & tHashOutput components store in cache memory enhances runtime performance.

Advantages of Batch Processing:

• Batch processing can efficiently handle large datasets.
• It takes minimal time to process the data even after data transformation.
• By grouping records from a large dataset and processing them as a single unit, it can be highly beneficial for improving performance.
• With the batch processing, it can easily scale to accommodate growing data volumes.
• It is particularly useful for operations like generating reports, performing data integration, and executing complex transformations on large datasets.

For more details: Get-started-talend-open-studio-data-integration

Note: Efficient Batch Processing in Talend-Part 2

In this episode of “What If? So What?”, Jim Hertzfeld sits down with Wally Burchfield, former senior executive at GM, Nissan, and Nissan United, to explore what’s driving transformation in the automotive industry and beyond. 

 Wally’s perspective is clear: in a world obsessed with automation and data, the companies that win will be the ones that stay human. 

 From “Build and Sell” to “Know and Serve” 

 The old model was simple: build a car, sell a car, repeat. But as Wally explains it, that formula no longer works in a world where customer expectations are shaped by digital platforms and instant personalization. “It’s not just about selling a product,” he said. “It’s about retaining the customer through a high-quality experience one that feels personal, respectful, and effortless.” Every interaction matters, and every brand is in the experience business. 

 Data Alone Doesn’t Build Loyalty – Trust Does 

 It’s true that organizations have more data than ever before. But as Wally points out, it’s not how much data you have, it’s what you do with it. The real differentiator is how responsibly, transparently, and effectively you use that data to improve the customer experience. 

 “You can have a truckload of data but if it doesn’t help you deliver value or build trust, it’s wasted,” Wally said. 

 When used carelessly, data can feel manipulative. When used well, it creates clarity, relevance, and long-term relationships. 

 AI Should Remove Friction, Not Feeling 

 Wally’s take on AI is refreshingly grounded. He sees it as a tool to reduce friction, not replace human connection. Whether it’s scheduling service appointments via SMS or filtering billions of digital signals, the best AI is invisible, working quietly in the background to make the customer feel understood. 

 Want to Win? Listen Better and Faster 

 At the end of the day, the brands that thrive won’t be the ones with the biggest data sets; they’re the ones that move fast, use data responsibly, and never lose sight of the customer at the center. 

 Listen to the full conversation with Wally Burchfield for more on how trust, data, and AI can work together to build lasting customer relationships—and why the best strategies are still the most human. 

Subscribe Where You Listen

Apple | Spotify | Amazon | Overcast | Watch the full video episode on YouTube

Meet our Guest – Wally Burchfield

Wally Burchfield is a veteran automotive executive with deep experience across retail, OEM operations, marketing, aftersales, dealer networks, and HR. 

He spent 20 years at General Motors before joining Nissan, where he held multiple VP roles across regional operations, aftersales, and HR. He later served as COO of Nissan United (TBWA), leading Tier 2/3 advertising and field marketing programs to support dealer and field team performance. Today, Wally runs a successful consulting practice helping OEMs, partners, and dealer groups solve complex challenges and drive results. A true “dealer guy”, he’s passionate about improving customer experience, strengthening OEM-dealer partnerships, and challenging the status quo to unlock growth. 

Follow Wally on LinkedIn  

Learn More about Wally Burchfield

Meet our Host

Jim Hertzfeld is Area Vice President, Strategy for Perficient.

For over two decades, he has worked with clients to convert market insights into real-world digital products and customer experiences that actually grow their business. More than just a strategist, Jim is a pragmatic rebel known for challenging the conventional and turning grand visions into actionable steps. His candid demeanor, sprinkled with a dose of cynical optimism, shapes a narrative that challenges and inspires listeners.

Connect with Jim:

LinkedIn | Perficient

Enterprise intelligence is redefining the future of business. For modern organizations the ability to harness information and turn it into strategic insight is no longer optional, it’s essential. Organizations are increasingly recognizing that enterprise intelligence is the catalyst for smarter decisions, accelerated innovation, and transformative customer experiences. As the pace of digital transformation quickens, those who invest in intelligent technologies are positioning themselves to lead.

IDC Market Glance: Enterprise Intelligence Services, 3Q25

We’re proud to share that Perficient has once again been included  in IDC’s Market Glance: Enterprise Intelligence Services report (doc #US52792625, September 2025). This marks our second consecutive year being included in the “IT Services Providers with Enterprise Intelligence Services offerings” category, which we believe reinforces our commitment to delivering innovative, data-driven solutions that empower enterprise transformation.

IDC defines Enterprise Intelligence as “as an organization’s capacity to learn combined with its ability to synthesize the information it needs in order to learn and to apply the resulting insights at scale by establishing a strong data culture.”

We believe our inclusion highlights Perficient’s continued investment in enterprise intelligence capabilities and our ability to embed these technologies into traditional IT services to drive smarter, faster outcomes for our clients.

We’re honored to be included alongside other providers and remain committed to helping organizations harness the power of enterprise intelligence to unlock new opportunities and accelerate growth.

Engineering the Future of Enterprise Intelligence

IDC notes: “Many IT services providers with heritage in systems integration, application development and management, and IT infrastructure services have practices focusing on technical advice, implementation and integration, management, and support of enterprise intelligence technology solutions.”

We don’t just deliver data services. We engineer intelligent ecosystems powered by AI that bring your data strategy to life and accelerate enterprise transformation. Our Data practice integrates every facet of enterprise intelligence, with a focus on AI-driven strategy, implementation, integration, and support of advanced, end-to-end technologies that reshape how businesses think, operate, and grow.

The future of enterprise intelligence is more than data collection. It’s about building adaptive, AI-enabled frameworks that learn, evolve, and empower smarter, faster decision-making.

To discover how Perficient can help you harness the power of enterprise intelligence and stay ahead of digital disruption, visit Data Solutions/Perficient.

Adverse drug reaction (ADR) detection is a primary regulatory and patient-safety priority for life sciences and health systems. Traditional pharmacovigilance methods often depend on delayed signal detection from siloed data sources and require extensive manual evidence collection. This legacy approach is time-consuming, increases the risk of patient harm, and creates significant regulatory friction. For solution architects and engineers in healthcare and finance, optimizing data infrastructure to meet these challenges is a critical objective and a real headache.

Combining the Databricks Lakehouse Platform with Agentic AI presents a transformative path forward. This approach enables a closed-loop pharmacovigilance system that detects high-quality safety signals in near-real time, autonomously collects corroborating evidence, and routes validated alerts to clinicians and safety teams with complete auditability. By unifying data and AI on a single platform through Unity Catalog, organizations can reduce time-to-signal, increase signal precision, and provide the comprehensive data lineage that regulators demand. This integrated model offers a clear advantage over fragmented data warehouses or generic cloud stacks.

The Challenges in Modern Pharmacovigilance

To build an effective pharmacovigilance system, engineers must integrate a wide variety of data types. This includes structured electronic health records (EHR) in formats like FHIR, unstructured clinical notes, insurance claims, device telemetry from wearables, lab results, genomics, and patient-reported outcomes. This process presents several technical hurdles:

• Data Heterogeneity and Velocity: The system must handle high-velocity streams from devices and patient apps alongside periodic updates from claims and EHR systems. Managing these disparate data types and speeds without creating bottlenecks is a significant challenge.
• Sparse and Noisy Signals: ADR mentions can be buried in unstructured notes, timestamps may conflict across sources, and confounding variables like comorbidities or polypharmacy can obscure true signals.
• Manual Evidence Collection: When a potential signal is flagged, safety teams often must manually re-query various systems and request patient charts, a process that delays signal confirmation and response.
• Regulatory Traceability: Every step, from detection to escalation, must be reproducible. This requires clear, auditable provenance for both the data and the models used in the analysis.

The Databricks and Agentic AI Workflow

An agentic AI framework running on the Databricks Lakehouse provides a structured, scalable solution to these problems. This system uses modular, autonomous agents that work together to implement a continuous pharmacovigilance workflow. Each agent has a specific function, from ingesting data to escalating validated signals.

Step 1: Ingest and Normalize Data

The foundation of the workflow is a unified data layer built on Delta Lake. Ingestion & Normalization Agents are responsible for continuously pulling data from various sources into the Lakehouse.

• Continuous Ingestion: Using Lakeflow Declarative Pipelines and Spark Structured Streaming, these agents ingest real-time data from EHRs (FHIR), claims, device telemetry, and patient reports. Data can be streamed from sources like Kafka or Azure Event Hubs directly into Delta tables.
• Data Normalization: As data is ingested, agents perform crucial normalization tasks. This includes mapping medical codes to standards like RxNorm, SNOMED, and LOINC. They also resolve patient identities across different datasets using both deterministic and probabilistic linking methods, creating a canonical event timeline for each patient. This unified view is essential for accurate signal detection.

Step 2: Detect Signals with Multimodal AI

Once the data is clean and unified, Signal Detection Agents apply a suite of advanced models to identify potential ADRs. This multimodal approach significantly improves precision.

• Multimodal Detectors: The system runs several types of detectors in parallel. Clinical Large Language Models (LLMs) and fine-tuned transformers extract relevant entities and context from unstructured clinical notes. Time-series anomaly detectors monitor device telemetry for unusual patterns, such as spikes in heart rate from a wearable.
• Causal Inference: To distinguish true causality from mere correlation, statistical and counterfactual causal engines analyze the data to assess the strength of the association between a drug and a potential adverse event.
• Scoring and Provenance: Each potential ADR is scored with an uncertainty estimate. Crucially, the system also attaches provenance pointers that link the signal back to the specific data and model version used for detection, ensuring full traceability.

Step 3: Collect Evidence Autonomously

When a candidate signal crosses a predefined confidence threshold, an Evidence Collection Agent is activated. This agent automates what is typically a manual and time-consuming process.

• Automated Assembly: The agent automatically assembles a complete evidence package. It extracts relevant sections from patient charts, re-runs queries for lab trends, fetches associated genomics variants, and pulls specific windows of device telemetry data.
• Targeted Data Pulls: If the initial evidence is incomplete, the agent can plan and execute targeted data pulls. For example, it could order a specific lab test, request a clinician chart review through an integrated system, or trigger a patient survey via a connected app to gather more information on symptoms and dosing adherence.

Step 4: Triage and Escalate Signals

With the evidence gathered, a Triage & Escalation Agent takes over. This agent applies business logic and risk models to determine the appropriate next step.

• Composite Scoring: The agent aggregates all collected evidence and computes a composite risk and confidence score for the signal. It applies configurable business rules based on factors like event severity and regulatory reporting timelines.
• Intelligent Escalation: For high-risk or ambiguous signals, the agent automatically escalates the issue to human safety teams by creating tickets in systems like Jira or ServiceNow. For clear, high-confidence signals that pose a lower operational risk, the system can be configured to auto-generate regulatory reports, such as 15-day expedited submissions, where permitted.

Step 5: Enable Continuous Learning

The final agent in the workflow closes the loop, ensuring the system improves over time. The Continuous Learning Agent uses feedback from human experts to refine the AI models.

• Feedback Integration: Outcomes from chart reviews, follow-up labs, and final regulatory adjudications are fed back into the system’s training pipelines.
• Model Retraining and Versioning: This new data is used to retrain and refine the signal detectors and causal models. MLflow tracks these updates, versioning the new models and linking them to the training data snapshot. This creates a fully auditable and continuously improving system that meets strict regulatory standards for model governance.

The Technical Architecture on Databricks

The power of this workflow comes from the tightly integrated components of the Databricks Lakehouse Platform.

• Data Layer: Delta Lake serves as the single source of truth, storing versioned tables for all data types. Unity Catalog manages fine-grained access policies, including row-level masking, to protect sensitive patient information.
• Continuous ETL & Feature Store: Delta Live Tables provide schema-aware pipelines for all data engineering tasks, while the integrated Feature Store offers managed feature views for models, ensuring consistency between training and inference.
• Detection & Inference: Databricks provides integrated GPU clusters for training and fine-tuning clinical LLMs and other complex models. MLflow tracks experiments, registers model versions, and manages deployment metadata.
• Agent Orchestration: Lakeflow Jobs coordinate the execution of all agent tasks, handling scheduling, retries, and dependencies. The agents themselves can be lightweight microservices or notebooks that interact with Databricks APIs.
• Serving & Integrations: The platform offers low-latency model serving endpoints for real-time scoring. It can integrate with clinician portals via SMART-on-FHIR, ticketing systems, and messaging services to facilitate human-in-the-loop workflows.

Why This Approach Outperforms Alternatives

Architectures centered on traditional data warehouses like Snowflake often struggle with this use case because they separate storage from heavy ML compute. Tasks like LLM inference and streaming feature engineering require external GPU clusters and complex orchestration, which introduces latency, increases operational overhead, and fractures data lineage across systems. Similarly, a generic cloud stack requires significant integration effort to achieve the same level of data and model governance.

The Databricks Lakehouse co-locates multimodal data, continuous pipelines, GPU-enabled model lifecycles, and governed orchestration on a single, unified platform. This integration dramatically reduces friction and provides a practical, auditable, and scalable path to real-time pharmacovigilance. For solution architects and engineers, this means a faster, more reliable way to unlock real-time insights from complex healthcare data, ultimately improving patient safety and ensuring regulatory compliance.

Conclusion

By harnessing Databricks’ unified Lakehouse architecture and agentic AI, organizations can transform pharmacovigilance from a reactive, manual process into a proactive, intelligent system. This workflow not only accelerates adverse drug reaction detection but also streamlines evidence collection and triage, empowering teams to respond swiftly and accurately. The platform’s end-to-end traceability, scalable automation, and robust data governance support stringent regulatory demands while driving operational efficiency. Ultimately, implementing this modern approach leads to better patient outcomes, reduced risk, and a future-ready foundation for safety monitoring in life sciences.

Perficient is a Databricks Elite Partner.  Contact us to learn more about how to empower your teams with the right tools, processes, and training to unlock your data’s full potential across your enterprise.

Precision therapeutics for rare diseases as well as complex oncology cases is an area that may benefit from Agentic AI Closed-Loop (AACL) systems to enable individualized treatment optimization — a continuous process of proposing, testing, and adapting therapies for a single patient (N-of-1 trials).

N-of-1 problems are not typical for either clinicians or data systems. Type 2 diabetes in the US is more of an N-of-3.8×10^7 problem, so we’re looking at a profoundly different category of scaling. This lower number is not easier, because it implies existing treatment protocols have not been successful. N-of-1 optimization can discover effective regimens rapidly, but only with a data system that can manage dense multimodal signals (omics, time-series biosensors, lab results), provide fast model iteration, incorporate clinician-in-the-loop safety controls, and ensure rigorous provenance. We also need to consider the heavy cognitive load the clinician will be under. While traditional data analytics and machine learning algorithms will still play a key role, Agentic AI support can be invaluable.

Agentic AI Closed-Loop systems are relatively new, so let’s look at what a system designed to support this architecture would look like from the ground up.

Data Platform

First, let’s define the foundation of what we are trying to build. We need a clinical system that can deliver reproducible results with full lineage and enable safe automation to augment clinical judgement. That’s a decent overview of any clinical data system, so I feel like we’re on solid ground. I would posit that individualized treatment optimizations need a reduced iteration time from the standard, just because the smaller N means we have moved farther from the SoC, so there will likely be more experiments. Further, these experiments will need more clever validations. Siloed and fragmented data stores, disconnected data, disjoint model operationalization and heavy ETL are non-starters based on our foundational assumptions. A data lakehouse is a more appropriate architecture.

A data lakehouse is a unified data architecture that blends the low-cost, flexible storage of a data lake with the structure and management capabilities of a data warehouse. This combined approach allows organizations to store and manage both structured and unstructured data types on cost-effective cloud storage, while also providing high-performance analytics, data governance, and support for ML and AI workloads on the same data. Databricks currently has the most mature lakehouse implementation. Databricks is well known for handling multimodal data, so the variety of data is not a problem even at high volume.

Clinical processes are heavily regulated. Fortunately, Unity Catalog provides a high level of security and governance across your data, ML, and AI artifacts. Databricks provides a platform that can deliver auditable, regulatory-grade systems in a much more efficient and effective way than siloed data warehouse or other cloud data stacks. Realistically, data provenance alone is not sufficient to align the clinician’s cognitive load with the smaller N; it’s still a very hard problem. Honestly, since we have had lakehouses for some time and have not been able to reliably tackle n-of-1 at scale, the problem can’t soly be with the data system. This is where Agentic AI enters the scene.

Agentic AI

Agentic AI refers to systems of autonomous agents, modular reasoning units that plan, execute, observe, and adapt, orchestrated to complete complex workflows. Architecturally, Agentic AI running on Databricks’ Lakehouse platform uniquely enables safe, scalable N-of-1 systems by co-locating multimodal data, high-throughput model training, low-latency inference, and auditable model governance. This architecture accelerates time-to-effective therapy, reduces clinician cognitive load, and preserves regulatory-grade provenance in ways that are materially harder to deliver on siloed data warehouses or generic cloud stacks. Here are some examples of components of the Agentic AI system that might be used as a foundation for building our N-of-1 therapeutics system. There can and will be more agents, but they will likely be used to enhance or support this basic set.

• Digital Twin Agents compile the patient’s multimodal state and historic responses.
• Planner/Policy Agents propose treatment variants (dose, schedule, combination) using constrained optimization informed by transfer learning from cohort data.
• Evaluation Agents collect outcome signals (biosensors, labs, imaging), compute reward/utility, and update the digital twin.
• Safety/Compliance Agents enforce clinical constraints, route proposals for clinician review when needed, and produce provenance records

For N-of-1 therapeutics, there are distinct advantages to designing agents to form a closed loop. Let’s discuss why.

Agentic AI Closed Loop System

Agentic AI Closed Loops (AACL)  enable AI systems to autonomously perceive, decide, act, and adapt within self-contained feedback cycles. The term “agentic” underscores the AI’s ability to proactively pursue goals without constant human oversight, while “closed loop” highlights its capacity to refine performance through internal feedback. This synergy empowers AACL systems to move beyond reactive processing, anticipating challenges and optimizing outcomes in real time. This is how we scale AI to realistically address clinician cognitive load within a highly regulated clinical framework.

• Perception: The AI gathers information from its from its Digital Twin among other sources.
• Reasoning and Planning: Based on its goals and perceived data of the current test iteration, the AI breaks down the objective into a sequence of actionable steps.
• Action: The AI executes its plan, often through the Planner/Policy Agents.
• Feedback and Learning: The system evaluates the outcome of its actions through the Evaluation Agents and compares them against its goals, referencing the Safety/Compliance Agents. It then learns from this feedback to refine its internal models and improve its performance in the next cycle. 

AAIC systems are modular frameworks. Let’s wrap up with a proposed reference architecture or an AAIC system using Databricks.

AAIC on Databricks

We’ll start with a practical implementation of the data layer. Delta Lake provides versioned tables for EHR (FHIR-parquet), structured labs, medication history, genomics variants, and treatment metadata. Time-series data like high-cardinality biosensor streams can be ingested via Spark Structured Streaming into using time-partitioning and compaction. Databricks Lakeflow is a solid tool for this. Patient and cohort embeddings can be stored as vector columns or integrated with a co-located vector index.

The Feature and ETL Layer builds on Lakeflow’s capabilities. A declarative syntax and a UI provide for a low-code solution for building continuous pipelines to normalize clinica code and compute rolling features like time-windowed response metrics. The Databricks Feature Store patterns enable reusable feature views for inputs and predictors.

Databricks provides distributed GPU clusters for the model and agent layer as well as access to foundational and custom AI model. Lakeflow Jobs orchestrate agent execution, coordinate microservices (consent UI, clinician portal, device provisioning), and manage retries.

MLFlow manages most of the heavy lifting for serving and integration. You can serve low latency policy and summarization endpoints while supporting canary deployments and A/B testing. The integration endpoints can supply secure APIs for EHR actionability (SMART on FIHR) and clinician dashboards. You can also ensure the system meets audit and governance standards and practices using the MLFlow model registry as well as Unity Catalog for data/model access control.

Conclusion

Agentic AI closed-loop systems on a Databricks lakehouse offer an auditable, scalable foundation for rapid N-of-1 treatment optimization in precision therapeutics—especially for rare disease and complex oncology—by co-locating multimodal clinical data (omics, biosensors, labs), distributed GPU training, low-latency serving, and model governance (MLflow, Unity Catalog). Implementing Digital Twin, Planner/Policy, Evaluation, and Safety agents in a closed-loop workflow shortens iteration time, reduces clinician cognitive load, and preserves provenance for regulatory compliance, while reusable feature/ETL patterns, time-series versioning (Delta Lake), and vector indexes enable robust validation and canary deployments. Start with a strong data layer, declarative pipelines, and modular agent orchestration, then iterate with clinician oversight and governance to responsibly scale individualized N-of-1 optimizations and accelerate patient-specific outcomes.

Perficient is a Databricks Elite Partner.  Contact us to learn more about how to empower your teams with the right tools, processes, and training to unlock your data’s full potential across your enterprise.

Here’s Why Perficient’s Elite Partner Status Will Make a Difference 

Databricks recently announced its acquisition of Tecton, a leading real-time feature store, to expand its AI agent capabilities. The move, covered by outlets including Reuters, InfoWorld, and AiTech365, enhances Databricks’ real-time data processing within Agent Bricks, enabling ultra-fast, contextual decision-making across use cases such as fraud detection, personalization, and risk scoring. 

What Does the Tecton Acquisition Mean for Databricks? 

The acquisition is significant because it closes one of the largest gaps in moving AI from experimentation to production at scale. Building on Tecton’s long-standing partnership and investor ties with Databricks—as well as shared clients like Coinbase—the deal strengthens Databricks’ market position and accelerates its ambition to lead enterprise AI infrastructure. 

Why it Matters: 

Real-Time AI Capabilities 

• Tecton’s feature platform enables sub-10 ms feature serving, allowing AI agents and applications to respond instantly with fresh data (fraud detection, chatbots, personalized recommendations). 

• By bringing this into Databricks, enterprises can now handle data prep, model training, and real-time inference all on one unified platform. 

Seamless Workflows from Raw Data to Production AI 

• Traditionally, teams stitched together multiple tools for data engineering, feature engineering, and deployment—causing delays and friction. 

• With Tecton integrated, Databricks now offers a streamlined path from raw data → features → models → production, reducing complexity and accelerating time-to-value. 

Enterprise-Grade Scale and Governance 

• Companies often struggle to reuse features across teams or maintain consistency between training and production. Tecton standardizes this process, improving accuracy and governance.

• Now, enterprises get that consistency inside the Databricks intelligence ecosystem—without juggling extra platforms. 

Ecosystem Strength 

• The acquisition expands Databricks’ capabilities into feature store and real-time AI territory. 

• For enterprises already on Databricks, it eliminates the need to adopt yet another specialized tool. 

Perficient: An Elite Partner in Real-Time AI Adoption 

As of March 17, 2025, Perficient achieved Databricks Elite Partner status, recognizing our deep expertise across data engineering, AI/ML, analytics, and governance. With more than 160 Databricks-certified consultants, we deliver end-to-end, mission-ready solutions—from lakehouse design to MLOps/LLMOps—across Azure, AWS, and GCP. 

Perficient was on the ground at the Databricks Data + AI World Tour 2025 in Dallas on September 4th, immediately following the Tecton announcement. 

“The acquisition of Tecton marks a turning point in how enterprises operationalize AI. By embedding real-time feature serving directly into the Databricks Intelligence Platform, organizations can build AI systems that not only learn but also react instantly to business-critical events,” said Nick Passero, director of Databricks at Perficient.

“As a Databricks Elite Partner, Perficient is uniquely positioned to help clients harness this capability—guiding them through integration, optimizing their architectures, and accelerating their path from experimentation to enterprise-scale AI.”    

Now with Tecton’s real-time feature services in the fold, Perficient is ideally positioned to help clients integrate, optimize, and scale—turning data intensity into AI agility. 

Three Ways Perficient Can Help Clients Harness Tecton-Powered AI 

1. Integration Best Practices: Perficient guides clients on embedding Tecton-powered real-time feature serving into their existing Databricks environments—ensuring workflows remain seamless, performant, and cost-effective. 

1. Accelerated Innovation with Confidence: With Tecton’s sub-10 ms latency, sub-100 ms freshness, and 99.99% uptime guarantees, Perficient helps clients deploy responsive agentic systems backed by enterprise-grade reliability. 

1. Comprehensive, Future-Ready AI Architectures: From strategy and design to deployment and governance, Perficient architects full-stack AI systems—including real-time feature stores, MLOps pipelines, AI agents, and dashboards—for use cases spanning personalization to risk mitigation. 

Build with Databricks + Perficient 

Databricks’ acquisition of Tecton is a milestone in the evolution of AI infrastructure. It empowers enterprises to build AI agents that think, react, and adapt in real time. Perficient, with our Elite Partner status, deep technical expertise, and strategic advisory capabilities, is ready to help organizations harness this shift and translate it into measurable business outcomes. 

Want to get started? Reach out to our Databricks experts to explore how real-time AI agents powered by Tecton can redefine your business. 

The headlines are troubling but predictable. The Trump administration will launch a program next year to find out how much money an artificial intelligence algorithm could save the federal government by denying care to Medicare patients. Meanwhile, a survey of physicians published by the American Medical Association in February found that 61% think AI is “increasing prior authorization denials, exacerbating avoidable patient harms and escalating unnecessary waste now and into the future.”

We’re witnessing the healthcare industry’s narrow vision of AI in action: algorithms designed to say “no” faster and more efficiently than ever before. But what if we’re missing the bigger opportunity?

The Current AI Problem: Built to Deny, Not to Help

The recent expansion of AI-powered prior authorization reveals a fundamental flaw in how we’re approaching healthcare technology. “The more expensive it is, the more likely it is to be denied,” said Jennifer Oliva, a professor at the Maurer School of Law at Indiana University-Bloomington, whose work focuses on AI regulation and health coverage.

This approach creates a vicious cycle: patients don’t understand their benefits, seek inappropriate or unnecessary care, trigger costly prior authorization processes, face denials, appeal those denials, and ultimately either give up or create even more administrative burden for everyone involved.

The human cost is real. Nearly three-quarters of respondents thought prior authorization was a “major” problem in a July poll published by KFF, and we’ve seen how public displeasure with insurance denials dominated the news in December, when the shooting death of UnitedHealthcare’s CEO led many to anoint his alleged killer as a folk hero.

A Better Vision: The AI Concierge Approach

What if instead of using AI to deny care more efficiently, we used it to help patients access the right care more effectively? This is where the AI Concierge concept transforms the entire equation.

An AI Concierge doesn’t wait for a claim to be submitted to make a decision. Instead, it proactively:

• Educates patients about their benefits before they need care
• Guides them to appropriate providers within their network
• Explains coverage limitations in plain language before appointments
• Suggests preventive alternatives that could avoid more expensive interventions
• Streamlines pre-authorization by ensuring patients have the right documentation upfront

The Quantified Business Case

The financial argument for AI Concierge services is compelling:

Star Ratings Revenue Impact: A half-star increase in Medicare Star Ratings is valued at approximately $500 per member. For a 75,000-member plan, that translates to $37.5 million in additional funding. An AI Concierge directly improves patient satisfaction scores that drive these ratings.

Operational Efficiency Gains: Healthcare providers implementing AI-powered patient engagement systems report 15-20% boosts in clinic revenue and 10-20% reductions in overall operational costs. Clinics using AI tools see 15-25% increases in patient retention rates.

Cost Avoidance Through Prevention: Utilizing AI to help patients access appropriate care could save up to 50% on treatment costs while improving health outcomes by up to 40%. This happens by preventing more expensive interventions through proper preventive care utilization.

The HEDIS Connection

HEDIS measures provide the perfect framework for demonstrating AI Concierge value. With 235 million people enrolled in plans that report HEDIS results, improving these scores directly impacts revenue through bonus payments and competitive positioning.

An AI Concierge naturally improves HEDIS performance in:

• Preventive Care Measures: Proactive guidance increases screening and immunization rates
• Care Gap Closure: Identifies and addresses gaps before they become expensive problems
• Patient Engagement: Improves medication adherence and chronic disease management

Beyond the Pilot Programs

While government initiatives like the WISeR pilot program focus on “Wasteful and Inappropriate Service Reduction” through AI-powered denials, forward-thinking healthcare organizations have an opportunity to differentiate themselves with AI-powered patient empowerment.

The math is simple: preventing a $50,000 hospitalization through proactive care coordination delivers better ROI than efficiently denying the claim after it’s submitted.

AI Healthcare Concierge Implementation Strategy

For healthcare leaders considering AI Concierge implementation:

• Phase 1: Deploy AI-powered benefit explanation tools that reduce call center volume and improve patient understanding
• Phase 2: Integrate predictive analytics to identify patients at risk for expensive interventions and guide them to preventive alternatives
• Phase 3: Expand to comprehensive care navigation that optimizes both patient outcomes and organizational performance

The Competitive Advantage

While competitors invest in AI to process denials faster, organizations implementing AI Concierge services are investing in:

• Member satisfaction and retention (15-25% improvement rates)
• Star rating improvements ($500 per member value per half-star)
• Operational cost reduction (10-20% typical savings)
• Revenue protection through better member experience

Conclusion: Choose Your AI Future

The current trajectory of AI in healthcare—focused on denial optimization—represents a massive missed opportunity. As one physician noted about the Medicare pilot: “I will always, always err on the side that doctors know what’s best for their patients.”

AI Healthcare Concierge services align with this principle by empowering both patients and providers with better information, earlier intervention, and more effective care coordination. The technology exists. The business case is proven. The patient need is urgent.

The question isn’t whether AI will transform healthcare—it’s whether we’ll use it to build walls or bridges between patients and the care they need.

The choice is ours. Let’s choose wisely.

At Perficient, we believe that championing innovation begins with the bold leaders who live it every day. Today, we’re proud to recognize Madhu Mohan Kommu, a key driver in our Databricks Center of Excellence (CoE), for being named a Databricks Champion, one of the most coveted recognitions in the Databricks ecosystem.

This honor represents more than technical mastery; it reflects strategic impact, thought leadership, and the power to drive transformation across industries through smart architecture and scalable data solutions. Achieving Databricks Champion status unlocks priority access to exclusive events, speaking engagements, and community collaboration. It’s a mark of excellence reserved for those shaping the future of data and AI, with Madhu as a stellar example.

The Journey Behind the Recognition

Earning Champion status was no small feat. Madhu’s five-month journey began with his first Databricks certification and culminated in a nomination based on real customer impact, platform leadership, and consistent contributions to Perficient’s Databricks CoE. The nomination process spotlighted Madhu’s technical depth, thought leadership, and innovation across enterprise engagements.

From Spark to Strategy: A Legacy of Impact

Since 2019, Madhu has led initiatives for our enterprise clients, delivering platform modernization, transformation frameworks, and cutting-edge data quality solutions. His expertise in the Spark Distributed Processing Framework, combined with deep knowledge of PySpark and Unity Catalog, has made him a cornerstone in delivering high-value, AI-powered outcomes across industries.

“Personally, it’s a proud and rewarding milestone I’ve always aspired to achieve. Professionally, it elevates my credibility and brings visibility to my work in the industry. Being recognized as a Champion validates years of dedication and impact.” – Madhu Mohan Kommu, Technical Architect

Strengthening Perficient’s Position

Madhu’s recognition significantly strengthens Perficient’s role as a strategic Databricks partner, expanding our influence across regions, deepening pre-sales and enablement capabilities, and empowering customer engagement at scale. His leadership amplifies our ability to serve clients with precision and purpose.

Looking Ahead: Agentic AI & Beyond

Next up? Madhu plans to lead Perficient’s charge in Agentic AI within Databricks pipelines, designing use cases that deliver measurable savings in time, cost, and process efficiency. These efforts will drive value for both existing and future clients, making AI innovation more accessible and impactful than ever.

Advice for Future Champions

Madhu’s advice for those on a similar path is to embrace continuous learning, collaborate across teams, and actively contribute to Perficient’s Databricks CoE.

What’s Hot in Databricks Innovation

From Lakehouse Federation to Mosaic AI and DBRX, Madhu stays at the forefront of game-changing trends. He sees these innovations not just as tools, but as catalysts for redefining business intelligence.

Madhu’s story is a powerful reflection of how Perficient continues to lead with purpose, vision, and excellence in the Databricks community.

Perficient + Databricks

Perficient is proud to be a trusted Databricks elite consulting partner with 100s of certified consultants. We specialize in delivering tailored data engineering, analytics, and AI solutions that unlock value and drive business transformation.

Learn more about our Databricks partnership.

We are proud to share that Perficient has been recognized among notable providers in The Commerce Services Landscape, Q3 2025, Forrester’s authoritative overview of 40 global providers authored by Principal Analyst Chuck Gahun. We believe this recognition highlights Perficient’s role as a systems integrator driving innovation across the commerce ecosystem.

Why This Recognition Matters

We believe Forrester’s inclusion reflects more than market presence. To us, it signals our strategic alignment with the future of enterprise commerce. As organizations shift from legacy platforms to intelligent AI-first ecosystems, Perficient is helping clients reimagine how value is created, sustained, and scaled.

Forrester asked each provider included in the Landscape to select the top business scenarios for which clients select them and from there determined which are the extended business scenarios that highlight differentiation among the providers. Perficient is shown in the report for having selected B2B2B commerce, B2B2C commerce, and Extended Reality and Augmented Reality Commerce as the top reasons clients work with us out of those extended use cases. Notably, Perficient was only one of three providers included in the Landscape to have selected Extended Reality and Augmented Reality Commerce. We believe this is a differentiator for us.

We were also listed with a focus on three industries:

• Manufacturing and production of consumer products
• Manufacturing and production of industrial products
• Retail

What This Means for Perficient and Our Clients

Several themes have emerged from within the commerce landscape that Perficient is poised to take by storm, especially as it relates to taking an AI-first approach to client’s challenges and goals.

AI-First Differentiation: There is a need for providers to move beyond buzzwords and clearly define what AI-first means. Perficient’s focus on operationalizing AI through proprietary intellectual property and composable architectures positions us to lead this shift.

Vertical-Specific Solutions: It is important to tailor AI offerings by industry. Our deep expertise and work across several industry verticals, notably commerce work for manufacturing and retail clients, reflects this strategic direction with solutions that drive real outcomes in product discovery, engagement, and experience.

A New Lens on Customer Journeys: Intelligent commerce is transforming the customer experience from AI-integrated search to immersive product journeys, and Perficient is building the infrastructure for what comes next.

Outcome-Based Engagements: As clients demand measurable impact, our ability to structure engagements around business outcomes powered by AI-driven insights sets us apart.

Embrace the New Era of AI-First Commerce

We believe Perficient’s inclusion in the Forrester landscape is more than recognition. For us, it is a signal to enterprise leaders that we are ready to help them transform legacy systems into intelligent platforms, activate AI across the full commerce lifecycle, and deliver personalized immersive experiences that drive growth.

Whether you are navigating B2B complexity, scaling retail innovation, or exploring extended reality commerce, Perficient is the partner to help you lead with confidence.

Ready to build what comes next in commerce? Let’s talk about how AI-first transformation can reshape your business.

Forrester does not endorse any company, product, brand, or service included in its research publications and does not advise any person to select the products or services of any company or brand based on the ratings included in such publications. Information is based on the best available resources. Opinions reflect judgment at the time and are subject to change. For more information, read about Forrester’s objectivity 

We’re proud to share that Perficient’s What If? So What? podcast has been named a Gold Stevie® Award winner in the Technology Podcast category at the 22nd Annual International Business Awards®. These awards are among the world’s top honors for business achievement, celebrating innovation, impact, and excellence across industries.

Winners were selected by more than 250 executives worldwide, whose feedback praised the podcast’s ability to translate complex digital trends into practical, high-impact strategies for business and technology leaders.

Hosted by Jim Hertzfeld, Perficient’s AVP of Strategy, the podcast explores the business impact of digital transformation, AI, and disruption. With guests like Mark Cuban, Neil Hoyne (Google), May Habib (WRITER), Brian Solis (ServiceNow), and Chris Duffey (Adobe), we dive into the possibilities of What If?, the practical impact of So What?, and the actions leaders can take with Now What?

The Stevie judges called out what makes the show stand out:

• “What If? So What? Podcast invites experts from different industries, which is important to make sure that audiences are listening and gaining valuable information.”
• “A sharp, forward-thinking podcast that effectively translates complex digital trends into actionable insights.”
• “With standout guests like Mark Cuban, Brian Solis, and Google’s Neil Hoyne, the podcast demonstrates exceptional reach, relevance, and editorial curation.”

In other words, we’re not just talking about technology for technology’s sake. We’re focused on real business impact, helping leaders make smarter, faster decisions in a rapidly changing digital world.

We’re honored by this recognition and grateful to our listeners, guests, and production team who make each episode possible.

If you haven’t tuned in yet, now’s the perfect time to hear why the judges called What If? So What? a “high-quality, future-forward show that raises the standard for business podcasts.”

Catch the latest episodes here: What If? So What? Podcast

Subscribe Where You Listen

APPLE PODCASTS | SPOTIFY | AMAZON MUSIC | OTHER PLATFORMS 

Watch Full Video Episodes on YouTube

Meet our Host

Jim Hertzfeld is Area Vice President, Strategy for Perficient.

For over two decades, he has worked with clients to convert market insights into real-world digital products and customer experiences that actually grow their business. More than just a strategist, Jim is a pragmatic rebel known for challenging the conventional and turning grand visions into actionable steps. His candid demeanor, sprinkled with a dose of cynical optimism, shapes a narrative that challenges and inspires listeners.

Connect with Jim: LinkedIn | Perficient