VORCL
← All courses
Beginner · 7 lessons

Your First AI Agent: From Zero to Autonomy

A course on what turns an LLM call into an agent: the decision loop, tools, memory, guardrails, and shipping to production via an API — from 50 lines of code to a GitHub issues triage agent.

Start →
  1. 01What Is an Agent: Loop, Tools, MemoryHow an agent differs from a chatbot and a workflow: the decide-act-observe loop, tools as the model's hands, and memory as its desk.35 min
  2. 02Your First Agent in 50 LinesWe build a minimal working agent in Python: a while loop, two tools, a call dispatcher, and honest stop conditions.45 min
  3. 03Tool Calling That WorksDesigning tool contracts: JSON Schema, descriptions, errors treated as results, parallel calls, and idempotency.45 min
  4. 04Memory and State: Short-Term, Long-Term, ScratchpadManaging context: trimming and summarizing history, a structured scratchpad for the plan, and external long-term memory with its pitfalls.45 min
  5. 05Guardrails: Budgets, Timeouts, and Human-in-the-LoopSetting boundaries for the agent: step, token, and money limits, tool timeouts, and mandatory human approval for irreversible actions.40 min
  6. 06The Agent as an API: Wrapping It in a ServiceTurning the script into a service: asynchronous runs instead of one long request, state storage, idempotency, and observability of every step.45 min
  7. 07Capstone: A GitHub Issues Triage AgentPutting it all together: the agent reads new issues, classifies them, suggests labels, waits for human approval, and keeps a quality report.60 min

Sources

Your First AI Agent: From Zero to Autonomy · VORCL