Advanced · ~2.7 hr · 8 lessons
Forward Deployed Engineer: Agents, Tools and Guardrails
Build bounded agent workflows in Python: validated tool calls, permissions, step limits, human approval, failure tests, guardrails and caching.
Enrollment is free. It unlocks every lesson in this course.
What's inside
- 1Workflow or Agent: Choosing Control Flow in CodeFree to read
- 2Tool Definitions and Validated Arguments in PythonEnroll to read
- 3Tool Permissions: The Application Enforces, Not the PromptEnroll to read
- 4Workflow Limits, Step Log and Resumable State in PythonEnroll to read
- 5Human Approval Gate for AI Agent Actions in PythonEnroll to read
- 6Failure Injection Tests for a Tool-Calling WorkflowEnroll to read
- 7An MCP Server for the Product CatalogEnroll to read
- 8LLM Output Guardrails and Tenant-Safe Response CachingEnroll to read
More in Forward Deployed Engineer
Intermediate · ~3.6 hr · 9 lessons
Forward Deployed Engineer: Python and FastAPI Foundations
Learn what a Forward Deployed Engineer does, then build a typed, tested FastAPI service in Python, explained for Java and Node engineers.
Intermediate · ~2.5 hr · 7 lessons
Forward Deployed Engineer: Build LLM Features
Build an LLM feature the way an engineer would: untrusted input, structured output, a provider adapter, streaming, one tool and a real-model lab.
Advanced · ~2.7 hr · 10 lessons
Forward Deployed Engineer: RAG, Search and Data
Build retrieval for enterprise documents: ingestion, chunking, keyword and vector search, access control, citations, evaluation and fresh data.
Advanced · ~1.4 hr · 5 lessons
Forward Deployed Engineer: Integrate and Ship
Place the AI service inside a customer's estate: pass identity, make calls reliable, write the design note and ship it into their environment.