5 hands-on AI projects that beat courses: RAG, coding agents, multi-agent and more
goyalshaliniuk · x · 2026-09-04
A "5 AI projects that teach you more than courses" roadmap, each with a concrete stack:
- RAG chatbot: embeddings, vector DBs, retrieval — Python + LangChain + Chroma + LLM API.
- AI coding agent: tool calling, memory, workflows — Python + OpenAI Agents SDK + GitHub API.
- Document Q&A: chunking & evaluation — Python + LlamaIndex + Qdrant + LLM.
- Multi-agent researcher: orchestration & planning — Python + CrewAI + web search + LLM.
The LangChain repo (146k stars) is attached as a starting reference.
More from coding & agent
- Software World: A Simulated GitHub Ecosystem Run by Agents for Multi-Agent Research — PangWeiKoh · 2026-09-04
- Developer Building 'Path of Codex', an ARPG Made with OpenAI's Codex — Dimillian · 2026-09-04
- GPT-6 demo shows fully autonomous 3D modeling in Blender, shipping scenes to UE5 — petewoodbridge · 2026-09-04
- Gemini CLI vs Crush vs Chrome DevTools MCP: three tools that aren't interchangeable — Different_Regret_146 · 2026-09-04
- OpenClaw's revamped onboarding lets users set up phone and Mac gateways easily — steipete · 2026-09-04
- Instinct launches interface-free personal agent living in iMessage and WhatsApp — Scobleizer · 2026-09-04