Databricks AI Engineer / Agent Bricks Engineer
Open a video or resource link, watch it, then check its box to track your progress and unlock what comes next.
- Build enterprise agents on Databricks with Agent Bricks, governed context and production app foundations
- Connect tools, orchestrate stateful workflows and ship usable agentic applications
- Evaluate, govern and optimize agent systems across quality, cost, security and operations
Benchmark: Databricks Applied AI Engineer + Agent Bricks / Genie Code / Data + AI platform engineering signals (current Sep 2026). Market refresh: Agent Bricks and custom agents; Databricks Apps; enterprise retrieval and context; MCP and governed tools; MLflow tracing/evals; Unity Catalog and Unity Gateway; state/memory; Genie Code and AI Runtime; production reliability and cost control.
Search Live Openings
We don’t have a direct job-board feed matched to this one yet, so here are real, live US job searches for “Databricks AI Engineer / Agent Bricks Engineer” instead.
Build enterprise agents on Databricks with Agent Bricks, governed context and production app foundations
Agent Bricks + custom agent foundations
Coverage: Agent Bricks / Custom Agents / Model Choice / Agent Harnesses / Databricks Apps / Python
Enterprise retrieval + AI-ready business context
Coverage: Mosaic AI Vector Search / RAG / Hybrid Retrieval / Reranking / Unity Catalog / Business Semantics
Connect tools, orchestrate stateful workflows and ship usable agentic applications
MCP, governed tools + external system integration
Coverage: MCP / Unity Catalog Functions / Tool Calling / External APIs / OAuth / Tool Permissions / Governance
Databricks Apps, state + agentic ML workflows
Coverage: Databricks Apps / Lakebase / Agent Memory / Genie Code / AI Runtime / Deployment / Stateful Agents
Evaluate, govern and optimize agent systems across quality, cost, security and operations
MLflow tracing, evals + continuous quality improvement
Coverage: MLflow 3 / Tracing / Agent Evaluation / LLM Judges / Prompt Registry / Production Monitoring
Unity Gateway governance + production economics
Coverage: Unity Gateway / Unity Catalog / Guardrails / Budgets / Routing / Observability / AI Runtime / Cost
Role-Readiness Proof / Capstone
Build a governed production agentic application on Databricks. Start with Agent Bricks or a custom agent and ground it in Unity Catalog data using high-quality retrieval; integrate at least two tools through MCP or governed functions; deploy a user-facing workflow with Databricks Apps and persistent state when needed; add a multi-agent branch only when it improves task quality; trace every critical step in MLflow, build an evaluation dataset with automated judges and regression gates; govern models and tools through Unity Gateway with permissions, budgets, guardrails and routing; profile quality, latency and cost, run a failure/security drill, and publish a runbook plus measured user and business outcomes.
Proof Rubric
- Prove reliable agent implementation
- authoritative enterprise context and retrieval quality
- clean MCP/tool contracts
- correct Unity Catalog permissions and Unity Gateway controls
- production Databricks Apps integration
- justified use of state and multi-agent orchestration
- MLflow traces and reproducible eval results
- cost/latency/quality discipline
- safe model and tool governance
- failure recovery and incident readiness
- measurable task success and adoption
- and clear tradeoffs among models, harnesses, managed vs custom agents, retrieval strategies, state stores and governance policies