Skip to main content

Vision

Agentic engineering is the discipline of building software systems where language models reason, call tools, retrieve context, ask for approval, and operate inside real product constraints.

What this playbook is​

This repository is a practical learning platform for developers who want to become production-ready AI Agentic Engineers. It is not a prompt collection and not a theory-first AI notes repo. The output is a public body of work: systems, diagrams, evaluations, demos, and deployable code.

Who it is for​

  • backend engineers moving into AI products
  • frontend engineers building copilot experiences
  • full-stack engineers shipping internal AI tools
  • founders validating agentic SaaS products
  • consultants building enterprise AI delivery capability

What you will build​

  • a provider gateway with structured outputs and streaming
  • a RAG copilot with citations and retrieval controls
  • a LangGraph-style orchestrator with tool execution and approvals
  • an MCP toolkit that exposes enterprise-safe tools
  • an Angular-based operator UI for agent state and approvals
  • production controls for evals, observability, CI/CD, and deployment

Product standard​

Every module should answer:

  • what problem are we solving?
  • what code would a real team ship?
  • how does it fail in production?
  • what artifact goes into the portfolio?
  • how could this become a product or service?

Success definition​

By the end of the playbook, a learner should be able to:

  • design an agentic system architecture
  • implement core flows in TypeScript and Python
  • explain tradeoffs around tools, context, retrieval, and approvals
  • instrument evals, costs, traces, and release quality
  • publish a serious GitHub portfolio that signals applied AI engineering ability