outpaczFlagship: AI Engineer Program
A 16-week hands-on program that takes you from GenAI basics to production RAG and agents, with 3 deployed projects, a capstone and interview prep.
51 lessons · 92 hours · ₹18,000
What you will build
Build with LLM APIs like an engineer
Async Python, Pydantic, structured outputs, prompt testing and a deployed LLM service
Ship production RAG
Ingestion, hybrid retrieval, reranking, text-to-SQL, multilingual RAG and evals
Engineer agentic systems
LangGraph, MCP, memory, human-in-the-loop and multi-agent designs
Operate AI in production
Evaluation CI, tracing, cost optimisation, AWS, queues and Kafka
Work with open models and safety
LoRA fine-tuning, guardrails, red-teaming and DPDP awareness
Be interview-ready
Three deployed gate projects, a capstone, a GitHub portfolio and mock interviews
Curriculum
Bridge videos (prerequisites)
- Python for LLM work · 180 min
- SQL and DBMS basics · 180 min
- APIs and auth · 120 min
Week 1: Dev setup and model landscape
- Recorded: LLM map, API anatomy, async Python, Pydantic and Git · 75 min
- Live build: Claude, OpenAI and Gemini calls and a model bake-off · 120 min
- Live lab: structured-output CLI with a Claude Code or Cursor workflow · 120 min
Week 2: Prompt engineering at engineering depth
- Recorded: prompt patterns, system prompts, few-shot, reasoning models and prompt versioning · 75 min
- Live build: prompt clinic in Workbench and AI Studio · 120 min
- Live lab: prompt test suite with promptfoo · 120 min
Week 3: Backend for AI apps
- Recorded: FastAPI, streaming, auth, Postgres and Redis · 75 min
- Live build: chat API with history in Postgres · 120 min
- Live lab: Dockerise and deploy to Render or AWS · 120 min
Week 4: Multimodal, automation and Gate 1
- Recorded: vision, PDF and audio with Whisper, function calling and n8n · 75 min
- Live build: voice and document assistant · 120 min
- Live lab: Gate 1 demo day, deployed LLM service · 120 min
Week 5: RAG foundations and ingestion
- Recorded: embeddings, chunking and parsing with Docling, Unstructured and LlamaParse · 75 min
- Live build: naive RAG on real PDFs · 120 min
- Live lab: ingestion lab on messy scans and tables · 120 min
Week 6: Retrieval quality
- Recorded: vector database selection, hybrid search, reranking and query transforms · 75 min
- Live build: hybrid search with reranking · 120 min
- Live lab: benchmark on 30 golden questions · 120 min
Week 7: Advanced RAG
- Recorded: text-to-SQL, GraphRAG, citations and multilingual RAG · 75 min
- Live build: documents plus SQL router · 120 min
- Live lab: Hindi and Hinglish RAG · 120 min
Week 8: RAG evaluation, security and Gate 2
- Recorded: Ragas, Langfuse, prompt injection and access control · 75 min
- Live build: evaluation harness · 120 min
- Live lab: Gate 2 demo day, production RAG with eval report · 120 min
Week 9: Agent fundamentals
- Recorded: workflows vs agents, ReAct and tool use · 75 min
- Live build: agent from scratch · 120 min
- Live lab: tool-design lab · 120 min
Week 10: LangGraph and agent SDKs
- Recorded: state graphs, persistence, human-in-the-loop and agent SDKs · 75 min
- Live build: LangGraph research agent · 120 min
- Live lab: human-approval flow · 120 min
Week 11: MCP, memory and context engineering
- Recorded: MCP servers, memory, context engineering, browser and computer-use agents · 75 min
- Live build: custom MCP server · 120 min
- Live lab: memory-enabled assistant · 120 min
Week 12: Multi-agent systems and Gate 3
- Recorded: supervisor and sub-agent patterns, cost vs accuracy · 75 min
- Live build: multi-agent app · 120 min
- Live lab: Gate 3 demo day, multi-agent app with MCP tools · 120 min
Week 13: LLMOps, evaluation and observability
- Recorded: eval-driven development, tracing and CI for prompts and agents · 75 min
- Live build: evaluation CI pipeline with GitHub Actions · 120 min
- Live lab: Langfuse tracing and dashboard · 120 min
Week 14: Scale and cloud
- Recorded: AWS (Lambda, ECS, S3, Bedrock), Kafka and queues, caching and model routing · 75 min
- Live build: queue-based document pipeline · 120 min
- Live lab: cost-optimisation lab · 120 min
Week 15: Open models, fine-tuning and safety
- Recorded: Ollama and vLLM, LoRA and PEFT, guardrails, red-teaming and DPDP · 75 min
- Live build: LoRA fine-tune of a small model on Colab · 120 min
- Live lab: red-team and guardrails lab · 120 min
Week 16: Capstone and career
- Recorded: system design for AI products, portfolio and interview prep · 75 min
- Live build: capstone clinic · 120 min
- Live lab: capstone demo day and mock interviews · 120 min
Syllabus
Who this program is for
Aspiring AI engineers who want a complete path from GenAI basics to production RAG and agents, with a portfolio to show employers.
Prerequisites
Python, APIs and DBMS. The bridge videos at the start cover all three, about 8 hours in total.
How the program runs
- Four phases: GenAI engineering foundations, RAG, agentic systems, and production AI with a capstone
- Three gate projects: each phase from 1 to 3 ends with a demo and code review by a senior mentor, with one resubmission allowed
- Recorded lessons are released every Monday
- Two live sessions each week: one for building and one for labs and doubt solving. Replays are available
- A problem statement or lab every week, reviewed by mentors, with a leaderboard
- Batch community group with daily doubt-desk hours
- Live session timings are announced before each batch starts
Time commitment
About 84 hours of core sessions over 16 weeks, plus 8 hours of bridge videos.
Assessment and certificate
- Weekly labs
- Three gate projects
- Capstone, with a final viva
- Verified certificate on scoring 70% or above with the capstone submitted
