Description:

Abstract

Join our engaging session to uncover the transformative power of Semantic Search and Conversational AI. In this session we will be presenting a combination of Semantic search and LLMs applied to information retrieval from pdf documents.
Semantic Search is a powerful tool that revolutionizes document retrieval in repositories through which we can achieve faster and relevant search results. By combining it with Langchain + LLMs we can unleash the capabilities of Conversational AI and build a chatbot that enables users to ask questions related to documents in the repository and provide comprehensive answers. It first retrieves relevant documents and then utilizes the power of LLM for intelligent responses. Through the harmonious integration of Semantic Search and Conversational AI, attendees will learn how to revolutionize their document management systems and elevate user interactions new heights. We will explore the synergy of these technologies and witness how they can enhance information retrieval process and human-machine conversations.

Flow of the talk:

1. Introduction to Semantic Search: Understanding the principles and advantages of semantic similarity-based document retrieval.
2. Understanding Vectors: Unraveling the mechanics of vectors and how it complements semantic search.
3. Harnessing the Power of LLM: Showcasing how LLM empowers the chatbot to provide accurate and contextually-aware answers.
4. Introduction to Conversational AI: Understanding the foundations of building interactive chatbots.
5. Document-centric Q&A: Explaining how the chatbot identifies related documents from the repository based on user questions.
6. Real-world Applications: Illustrating hybrid approach offers tangible benefits in various domains.
7. Python Implementation: Demonstrating practical implementation and the step-by-step development of the chatbot using Langchain and Python.
8. Real-world Impact: Presenting use cases where this conversational AI can bring efficiency and convenience to knowledge retrieval.

Prerequisites:

A basic understanding of the following is good to have: - Python - Machine Learning - LLMs

Speaker Info:

Rushiraj Chavan, a Data Scientist at Gramener, specializes in machine learning, computer vision, and deep learning. He has extensive experience in projects like Supply Chain optimization, Generative AI, Computer Vision, and Audio Signal Processing, reflecting his profound understanding of complex project implementations. Staying up-to-date with the latest developments in the field, Rushiraj is dedicated to exploring new techniques and models.

Anupama Deo is an Associate Data Scientist at Gramener and brings deep expertise in the fields of ML, NLP, Deep Learning, and Gen AI. She also has hands-on expertise in Voice Cloning techniques and Audio Signal Processing. With a strong background in these areas, Anupama brings valuable insights to her projects and stays committed to continuous learning and growth in the field.

Speaker Links:

Speaker 1:

LinkedIn: https://www.linkedin.com/in/rushirajchavan/

Spoke about Stable Diffusion at PyCon Limerick: https://pycon.ie/previous-pycons/pycon-limerick-2023/

Speaker 2:

Linkedin: https://www.linkedin.com/in/anupama-deo/