Webinar | RAG, Built Live: From Concept to Deployed App with Genspark

About

Most RAG demos prove nothing. Ask a chatbot about the "Attention Is All You Need" paper — with or without retrieval — and it answers perfectly. So how do you know RAG is actually doing anything? In this 90-minute hands-on build-along, we'll fix that. We'll build two working RAG applications, live, over a corpus deliberately designed to break every model without retrieval. What we'll build together (on your own laptop): 🟦 Track A — No-code RAG in 15 minutes A grounded Custom Agent on Genspark that answers only from your documents, cites sources, and refuses when it doesn't know. You'll leave with a shareable link to something that works. 🟥 Track B — The real pipeline, deployed Genspark Code builds and deploys a full RAG application: chunking, embeddings, vector index, hybrid retrieval, reranking, grounded generation with citations. You'll see every component you just heard about in the theory section actually get built. What you'll walk out with: A deployed RAG app you can point at your own documents A 15-document corpus pack to test and keep experimenting with Copy-paste prompt kit for both tracks A mental model of RAG that survives contact with production The one diagnostic question that separates real RAG from theatre Session flow (90 min) 0:00 — What is RAG, and how do we know it's working? 0:15 — The full pipeline: chunking, embeddings, retrieval, reranking, grounding 0:30 — Track A: build a no-code RAG agent, live 0:50 — Break it on purpose: baseline vs RAG, 8 test questions 1:00 — Track B: build & deploy the real pipeline 1:20 — Evaluation, hardening, and when RAG is the wrong answer 1:26 — Share what you built + Q&A Who this is for Mixed audience by design. Business & product folks who want to understand what RAG really is beyond the buzzword. Engineers and data scientists who want a clean production mental model and a working reference app. Every technical concept has a plain-language version first. Prerequisites A free Genspark account (genspark.ai) A laptop with a stable connection and a second browser tab No prior RAG, coding, or ML experience required Curiosity encouraged, cynicism welcome About your host Abhishek Kumar — Enterprise AI & Data Science Leader, Genspark Genius (Bangalore Chapter). Ships production GenAI systems for a living, teaches on the side. Hosted by Genspark Genius Program · Bangalore Chapter Global community of practitioners running local AI workshops, hackathons, and build sessions. Session will be recorded. Attendees receive the recording, the deck, the facilitator handout, and the corpus pack within 24 hours.