How Addepar Scales Investment Workflows with Databricks AI Agents

2026-04-02 · Source: Databricks · Field: Finance & Economics — Capital Markets & Investment Management, FinTech & Digital Financial Services · Depth: Intermediate, short

Summary

Addepar has transformed its infrastructure on Databricks to develop Addison, a safe and scalable Generative AI system for financial services. This architecture prioritizes security, data privacy, and governance, integrating Unity Catalog for permissions and access controls. Addepar hosts frontier models within its environment via Databricks Model Serving and manages them with MLflow, ensuring consistent lifecycle management and auditability. The system keeps personally identifiable and client-identifiable data within the Addepar ecosystem to meet compliance and jurisdictional requirements. Addepar is also adopting Databricks Agent Bricks to evolve from simple LLM prompts to agentic workflows, using Supervisor Agents to coordinate Genie-powered analytics and execute multi-step actions. This unified platform approach, leveraging Databricks Foundation Model APIs and Managed MLflow, centralizes agent workflow definition, testing, and deployment with integrated governance, enhancing collaboration and data sharing internally and with clients.

Key takeaway

For CTOs and VP of Engineering in financial services building GenAI applications, adopting a unified data and AI platform like Databricks is crucial. This approach allows you to maintain stringent data governance, ensure auditability, and manage agentic workflows within a single, secure environment, significantly simplifying compliance and accelerating the deployment of reliable AI solutions for your clients.

Key insights

Unified data, governance, and agent orchestration are critical for secure, scalable GenAI in regulated financial services.

Principles

• Data and models must remain within the ecosystem.
• Governance guarantees should extend to AI systems.
• Agentic workflows enhance reliability for production.

Method

Addepar uses Databricks Unity Catalog for access control, Model Serving for hosting LLMs, MLflow for lifecycle management, and Agent Bricks for orchestrating agentic workflows, all within a unified platform.

In practice

• Integrate data governance with AI stack.
• Use agent frameworks for complex workflows.
• Centralize model and data collaboration.

Topics

• Databricks AI Agents
• Financial Services AI
• Data Governance
• Agent Bricks
• MLflow

Best for: CTO, VP of Engineering/Data, AI Architect, AI Engineer, MLOps Engineer, Director of AI/ML

Related on AIssential

• Agent Bricks: The Governed Enterprise Agent Platform — Enterprise AI agents require a platform approach for deep business integration, governance, and contextual accuracy.
• Introducing Genie One, Genie Agents, and Genie Ontology — Databricks' Genie suite provides context-aware AI agents for enterprise data, significantly improving accuracy and speed.
• Databricks says it solved the decades-old data pipeline problem that's been slowing AI agents — Databricks unifies transactional and analytical data at the storage layer, enabling millisecond latency for AI agents.
• Why Financial AI Can't Scale Without Unified Governance with James Dean of Google and Mark Crean of Securiti — Effective data governance, not technology, is the primary bottleneck for scaling AI in financial services.
• The AGI moment? Databricks’ new releases zero in on support and deployment of AI agents — Databricks' new platform architecture and tools aim to integrate AI agents with enterprise data for AGI, addressing context, cost, and control.
• The Agentic Enterprise Just Got Real — AI's primary challenge is context management, not intelligence, requiring robust governance for agentic systems.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Databricks.