- Dev fundamentalsWeek P1–P2
- Terminal, Git/GitHub, Python, JS/TS, React, HTTP, SQL, Postgres
- Engineering toolkitWeek 1
- uv, ruff, pytest, Pydantic, asyncio, tenacity, NumPy, pandas, SQLAlchemy, FastAPI, Docker, Redis
- AI coding toolsWeek 2
- Claude Code, Cursor, Windsurf, Copilot, Codex, Gemini CLI, Aider, v0, Lovable, Bolt, CLAUDE.md/AGENTS.md, worktrees
- Full-stackWeek 2, 4, 8
- Next.js App Router, Tailwind, shadcn/ui, Drizzle/Prisma, Vercel AI SDK
- ML foundationsWeek 3
- Vectors, softmax, train/test, metrics, scikit-learn, neural nets, backprop, PyTorch
- LLM internalsWeek 3
- BPE tokenization, embeddings, RoPE, attention, decoder-only, pre-training, SFT, RLHF, DPO, reasoning models, sampling, KV cache, MoE
- Models and APIsWeek 3
- Claude, GPT, Gemini, Llama, Qwen, DeepSeek, Sarvam; SDKs, OpenRouter, LiteLLM, Bedrock, Vertex, Batch APIs
- Open models and inferenceWeek 3
- Hugging Face, Ollama, llama.cpp, quantization (GGUF/AWQ/GPTQ), vLLM, SGLang, TGI, Groq, Modal, RunPod
- Prompt and context engineeringWeek 4
- System prompts, few-shot, XML, chain-of-thought, chaining, prompt caching, compaction, promptfoo, memory (Mem0, Zep, Letta)
- Structured output and toolsWeek 5
- JSON schema, Instructor, Outlines, BAML, function calling, tool design, text-to-SQL
- LLM frameworksWeek 5
- LangChain, LlamaIndex, Pydantic AI, Haystack
- Embeddings and vector DBsWeek 6
- OpenAI/Voyage/Cohere/BGE embeddings, pgvector, Pinecone, Qdrant, Weaviate, Chroma, Milvus, FAISS, HNSW
- RAGWeek 6
- Parsing (Docling, Unstructured, LlamaParse), crawling (Firecrawl), chunking, BM25, hybrid search, reranking, citations
- Advanced RAGWeek 7
- HyDE, query rewriting, contextual retrieval, ColBERT, ColPali, GraphRAG, Neo4j, agentic RAG
- EvalsWeek 7
- Error analysis, golden sets, LLM-as-judge, RAGAS, DeepEval, Braintrust, LangSmith, Langfuse, Phoenix
- OptimisationWeek 7
- DSPy (MIPROv2, GEPA), semantic caching
- SaaS engineeringWeek 8
- Auth (Clerk, Supabase), queues (Celery, Inngest), S3/R2, Razorpay, Stripe, metering, generative UI
- AgentsWeek 9
- ReAct, plan-and-execute, reflection, agent loop from scratch, budgets, memory
- Agent frameworksWeek 9
- LangGraph (state, checkpointers, interrupts, subgraphs, Studio), OpenAI Agents SDK, Claude Agent SDK, CrewAI, AutoGen, Google ADK, smolagents, Mastra
- MCP and multi-agentWeek 10
- MCP primitives and transports, FastMCP, OAuth, A2A, supervisor/swarm, subagents
- Autonomous agentsWeek 10
- Browser Use, Playwright, Stagehand, computer use, E2B sandboxes, Temporal, trajectory evals
- Multimodal and voiceWeek 11
- VLMs, OCR, CLIP, Whisper, Deepgram, ElevenLabs, Realtime APIs, LiveKit, Pipecat, Bhashini
- Fine-tuningWeek 11
- LoRA, QLoRA, SFT, DPO, GRPO, distillation, TRL, PEFT, Unsloth, Axolotl
- Generative mediaWeek 11
- Diffusion, Flux, Stable Diffusion, ComfyUI, Veo, Runway
- Production and LLMOpsWeek 12
- OpenTelemetry, Helicone, guardrails, Llama Guard, Presidio, red teaming, gateways, load testing, Kubernetes, GitHub Actions, Terraform
- Responsible AI and lawWeek 12
- OWASP LLM Top 10, DPDP Act, EU AI Act, bias, model licences
- Product and founder skillsWeek 13–14
- Validation, MVP scoping, AI system design, pricing, launch, SEO/GEO, analytics, funding
- Career skillsWeek 15
- Resume, portfolio, interviews, job platforms, negotiation