Build a Mini RAG App in 2 Hours — Live Demo with LangChain, Chroma & LangSmith

What we’ll do
A fast-paced, live demo where we build a tiny Retrieval-Augmented Generation (RAG) app end-to-end. We’ll keep installs to a minimum (think Google Colab) and focus on the “aha!” moments—how retrieval changes answers, how to trace runs, and what to tweak for quality.

Who this is for
Developers, data folks, solution architects, and curious PMs who want a practical, vendor-agnostic intro to RAG without a heavy setup.

You’ll see (live):

• Ingest a small doc set and create embeddings (Chroma)
• Wire up retrieval + generation with LangChain
• Compare single-query vs. multi-query and sub-query (“drill-down”) retrieval
• Use LangSmith to trace, compare, and reason about answer quality

You’ll take away:

• A handout with step-by-step instructions
• 4 demo notebooks + a tiny sample dataset
• A simple checklist for moving from demo to a real pilot
• Post-event WhatsApp support for Q&A while you practice

Agenda (120 minutes)

• 0:00–0:10 — RAG concepts, what we’re building
• 0:10–0:35 — Mini RAG app (end-to-end) + tracing
• 0:35–1:05 — Multi-query retrieval (when and why)
• 1:05–1:25 — Sub-query / drill-down retrieval
• 1:25–1:40 — Step-back/generification (rapid demo)
• 1:40–2:00 — Wrap-up, next steps, resources, Q&A

Before you join

• No deep ML background needed.
• Optional: a Google account for Colab if you want to follow along visually.
• We’ll share all materials after the session.

After the event
Have questions while practicing? Use our WhatsApp group and we’ll help during business hours.

Format: Live online (meeting link will be shared after RSVP)
Cost: Free
Level: Beginner–Intermediate

RSVP to save your spot!