Sponsored by: KPMG | Activation of Agentic AI in Data Engineering Leveraging Databricks Agent Bricks
Overview
| Experience | In Person |
|---|---|
| Track | Artificial Intelligence & Agents |
| Industry | Enterprise Technology, Healthcare & Life Sciences, Financial Services |
| Technologies | Unity Catalog, Databricks Agents, Lakebase |
| Skill Level | Intermediate |
As organizations accelerate their AI and data strategies, traditional data engineering faces growing challenges around scale, complexity and the need for real-time insights. This session will explore how agentic AI, powered by Databricks Agent Bricks, is redefining modern data engineering through autonomous, goal-driven agents that streamline data ingestion, transformation, quality management and orchestration across the data lifecycle. You’ll learn how Agent Bricks activates intelligent agents within the Databricks ecosystem to automate complex data workflows, strengthen governance and accelerate the delivery of trusted, AI-ready data products at enterprise scale.
Session Speakers
Bindu Birur
/Head of Data Eng & Analytics Delivery
KPMG LLP (US)
Sameer Bhagi
/Sr Engineer
KPMG
Full Summary
Inside Nexa: how KPMG built an agentic AI platform for the data development lifecycle
KPMG's Digital Nexus team set out to compress a fragmented, handoff-heavy data development lifecycle into one governed, AI-orchestrated flow. The result is Nexa, an internal platform in production that applies persona-driven agents across requirements, modeling, code generation, QA, and ongoing business validation. Built on the Databricks Platform, Nexa reflects a shift from coding copilots to end-to-end agentic systems that keep humans in the loop and preserve context from start to finish. Teams operating at this scale increasingly rely on data engineering infrastructure to unify pipelines, governance, and AI execution in a single managed environment.
FAQ
Nexa is an internal KPMG product built by the Digital Nexus team and powered by Databricks. It is used on KPMG delivery programs and is not sold as a standalone product.
No. Nexa automates routine transformation tasks while keeping humans in the loop for review, approval, and edits. Engineers shift from boilerplate to validating AI-generated artifacts and handling higher judgment work.
Nexa focuses on the transformation tier. Interpretation of requirements, code generation, unit tests, and business validation sit inside Nexa, while extraction and load continue to use KPMG's existing data movement frameworks.
Digitized requirements live beside the data, so business validations can run after every ETL load. Violations are flagged with severity, and the platform can stop downstream delivery before issues reach the client.
Personas map to security groups governed by Unity Catalog and role-based access control. Users see which agents they can run, what they can approve, and which artifacts they can edit inside the platform.