Hello Everyone,

We’re excited to announce a hands-on workshop at The Fifth Elephant 2025 for developers, AI practitioners, and researchers working in GenAI space (specially dealing with infromation extractions).

🔍 Workshop: Develop Visual Augmented Q&A using Multi-vector Vision Embeddings with Vector DB
🗓️ Duration: 4 hours
📍 Format: In-person + livestream for Fifth Elephant members

Why attend?
Documents today aren’t just text — they’re packed with infographics, charts, and images. In this advanced workshop, you’ll learn to build Vision-RAG systems that process both text and visual data, using Vision Language Models like ColPali and vector databases for complex document Q&A workflows.

What you’ll learn:

Build a Vision-RAG system for Visual Augmented Q&A
Work with multi-vector vision embeddings and Vector DBs
Integrate ColPali-based re-ranking and retrieval
Address practical challenges for production-ready Vision-RAG setups

Who should attend:

AI/ML Engineers, MLOps, Data Scientists
Researchers in GenAI and multi-modal AI
Product engineers building document or knowledge retrieval systems

Note: This is an intermediate level workshop.

Participants must be python friendly coders.
Participants should have basic working knowledge of conventional text-RAG systems !!
Beginners are welcome but expected to catch up on Python, LLM fundamentals, embeddings, and multi-modal AI basics.

📚 Workshop materials: GitHub Repository

🎟️ Reserve your seat:
Seats are limited and open on a first-come-first-serve basis for The Fifth Elephant 2025 ticket holders and members.
For questions, call us at +91-7676332020 or write to info@hasgeek.com

See you at the workshop!

Special note:
Many of you might have attended my workshop at The Fifth Elephant 2024 with Subhom Mitra — Turning Tables: Turn English to Data Using Self-Hosted LLMs. That workshop lit a spark around the idea of developing low-code/no-code solutions using code LLMs in the data space.

This year, I am throwing light into advancements happening in multi-modal AI space.

Seeing is retrieving: search chatbots with vision embeddings

{{ gettext('Draft') }}