Custom Agents now available on Databricks

2026-02-18 · Source: Databricks · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Cloud Computing & IT Infrastructure · Depth: Intermediate, quick

Summary

Databricks has released Custom Agents, formerly Agent Framework, enabling developers to build, test, and deploy production-quality AI agents as fully managed Databricks Apps. This offering allows teams to use existing tools and workflows, eliminating the need for code re-architecture or infrastructure management. Custom Agents include prebuilt skills and templates, integrated evaluation, and CI/CD pipeline integration to accelerate development from prototype to production. Agents run on serverless compute with built-in security and governance, featuring Lakebase-powered memory for context awareness and direct connectivity to enterprise data, all under unified governance. This aims to reduce custom integration and facilitate the deployment of trusted, domain-aware agents.

Key takeaway

For AI Architects and VP of Engineering evaluating platforms for AI agent development, Databricks Custom Agents offer a compelling solution by providing a fully managed, serverless environment with integrated governance and memory. This approach simplifies deployment and reduces operational overhead, allowing your teams to focus on agent logic and quality rather than infrastructure management. Consider leveraging its built-in features to accelerate your agent development lifecycle and ensure consistent data and model governance.

Key insights

Databricks Custom Agents streamline AI agent development and deployment with serverless infrastructure and integrated governance.

Principles

• Leverage existing tools and workflows.
• Integrate evaluation and CI/CD for quality.
• Ensure consistent governance across data and agents.

Method

Build agents locally, iterate with feedback loops, use prebuilt skills/templates, integrate into CI/CD, and deploy as serverless Databricks Apps with Lakebase memory.

In practice

• Utilize prebuilt agent skills for common tasks.
• Connect agents directly to enterprise data.
• Deploy agents as serverless Databricks Apps.

Topics

• AI Agents
• Databricks
• Serverless Compute
• MLOps
• Generative AI

Code references

• databricks/app-templates

Best for: AI Architect, CTO, VP of Engineering/Data, Machine Learning Engineer, AI Engineer, Software Engineer

Related on AIssential

• The Agentic Enterprise Just Got Real — AI's primary challenge is context management, not intelligence, requiring robust governance for agentic systems.
• Databricks partners with OpenAI on GPT-5.5 — GPT-5.5 significantly advances AI agent capabilities for complex enterprise tasks and coding workflows.
• AI/BI Genie, Foundational Model API, and Databricks Assistant Now Generally Available in AWS GovCloud — Databricks now offers advanced generative AI and data intelligence features within AWS GovCloud for regulated U.S. government use.
• Databricks nears $5bn equity raise at $134bn valuation — Databricks secured $7B in funding at a $134B valuation to expand its AI-driven data platform offerings.
• Accenture partners with Databricks on scaling enterprise AI solutions — Accenture and Databricks partnered to help enterprises scale AI by integrating data and deploying advanced AI agents.
• 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.

Open in AIssential →

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