Demo Proposal: Thursday 5 March - RAG System, Best approach (Hybrid Retrieval)

[murilofarias10]

• City Chapter: Vancouver
• Draft Title: RAG System, Best approach (Hybrid Retrieval)
• Length: 20 minutes

In this demo, I show how I built a Retrieval-Augmented Generation (RAG) I explain how I extract text from PDFs split the content into chunks, generate embeddings with OpenAI, and store them in FAISS for similarity search. Then, I show how the system retrieves the most relevant chunks and uses an LLM to generate accurate answers. I also discuss challenges like “lost in the middle” and how better retrieval improves response quality.

Speaker Bio (optional)

• Name: Murilo Farias
• Mini-bio: Data Engineering student based in Vancouver from Brazil with +7 years of experience.
• Would you like help with your demo? Either if you are a begginer or expert this is a deep understandable process that could be usefull daily.
• Social media link(s): linkedin: https://www.linkedin.com/in/murilofarias10/
  github: https://github.com/murilofarias10
  personal portfolio: https://murilofarias.netlify.app/

---

👍 Audience: React with 👍 on this issue to upvote this talk! Top-voted proposals get priority scheduling.

[VeVarunSharma]

hi @murilofarias10 , thank you for taking the time to write and share this! 🙏🏽

We're looking for talks more so around the GH platform, agentic DevOps, and general AI assisted development use cases and advanced methods. For a RAG showcase that would fit better for an azure user group meet up for example that would show case how to do with AI Foundry (Example).

I would suggest thinking around those concepts :) or think outside of the box too!