The Mother of AI Project - Phase 1: RAG Systems

This is a 6-week, hands-on series where you’ll build a complete, production-grade Retrieval-Augmented Generation (RAG) system - just like we do in the real world.

You’ll go from scratch to a fully functional AI Research Assistant that downloads data, chunks content, embeds, indexes, and answers questions - powered by OpenSearch 2.19, FastAPI, Ollama, Airflow 3.0, and more.

Cohort Duration

Start Date: August 10, 2025

End Date: September 14, 2025

Live Blog + Code Release: Every Thursday through Jam With AI Substack newsletter

Code walkthrough + Q&A Hours: Every Sunday, 16:00 - 17:30 CET

All sessions will be recorded and available if you miss them.

Week 1: FastAPI + Docker setup

📍 Code walkthrough + Q&A Hours (Aug 10, 16:00–17:30 CET): Architecture setup, Docker Compose, FastAPI bootstrapping

Week 2: arXiv scraping + parsing with Airflow

📍 Code walkthrough + Q&A Hours (Aug 17, 16:00–17:30 CET):

DAGs, parsing with GROBID & Docling, fallback logic

Week 3: Hybrid Search infra with OpenSearch

📍Code walkthrough + Q&A Hours (Aug 24, 16:00–17:30 CET):

Schema design, BM25 tuning, semantic vector search

Week 4: Chunking + Retrieval evaluation

📍 Code walkthrough + Q&A Hours (Aug 31, 16:00–17:30 CET):

Chunking techniques, precision-recall tradeoffs, evaluation metrics

Week 5: Full RAG pipeline + prompts

📍 Code walkthrough + Q&A Hours (Sept 7, 16:00–17:30 CET):

Prompt design, truncation strategy, LlamaIndex orchestration

Week 6: Langfuse Observability + Gradio UI

📍 Code walkthrough + Q&A Hours (Sept 14, 16:00–17:30 CET):

Monitoring, prompt versioning, UI setup, cache

We’re excited to have you on board, get ready to dive deep into building real-world AI systems with us! 💪

Jam With AI!