Introduction to RAG (Retrieval Augmented Generation)

The talk will cover ways to unlock the power of large language models (LLMs) through effective augmentation of prompts with relevant context, at run-time. We will explore ways of different approaches and sophistications possible within them.

The talk is for the intermediate level AI developers with basic background in Machine Learning, Deep Learning, etc. Specifically, to grasp the content, I recommend having a decent understanding of the following topics::
➡ Prompt Engineering
➡ Vector Embeddings
➡ Natural Language Processing, in general

Introduction to Graph Data Science

Typical data science is done on structured data such as tables, sometimes on unstructured data such as text, but graph is a bit different than both, unstructured as well as variable size. This talk gives overview of the challenges and how to apply Data Science on Graphs. It will start with introduction to graphs, then its challenges then will enumerate graph algorithms. At the end it will take up a few applications and then conclude with the next steps. The talk will also mention the practical usage using Neo4j platform.

Introduction to Autonomous Agents

Frameworks like Microsoft's Autogen facilitate building multiple agents to get tasks done. This session will introduce Autogen with an example.

Computing Midcurve of a Thin Polygon for Mechanical Engineering

Calculating physical product design is a complicated problem. Existing models provide rough examples in reasonable times but still require weeks of manual correcting.

This presentation explores how graphs can help mechanical engineering by modeling geometric polygonal figures into a graph representation. Neural networks can then train and learn from examples provided to compute the midcurve of a thin polygon, becoming a graph-to-graph operation akin to an encoder-decoder problem.

Implementing Generative AI FAQ Bot on own documents

In a world where technology continues to push boundaries, a remarkable feat is emerging at the intersection of AI and conversation. Imagine building a Google Bard or ChatGPT-style chatbot that not only mimics human interaction but responds using your own data. Intrigued? Let’s delve into a transformative workflow that brings this innovation to life. A bot making a Bot with Streamlit, Langchain, HuggingFace and VertexAI Palm APIs

Intro to Fine-tuning LLMs

The talk will cover ways to adapt the power of large language models (LLMs) to your own custom dataset or corpus. We will explore the need for fine-tuning, different approaches, and the advantages and disadvantages of each approach.

Audience needs decent background in Machine Learning, Deep Learning, Natural Language Processing, Large Language Models and Generative AI, in general.

(Mid) Career Transition to Artificial Intelligence

Topics:
➡️ Muscle Automation and Brain Automation 🧠
➡️ What AI and ML are, and what they're not 🤖
➡️ The 3 personas: user, developer, and researcher 👩‍💻👨‍🔬
I'll also discuss learning paths for each persona. 🧭

Need to have background in college level mathematics

Introduction to Generative AI

From Zero to Gen AI, talks about the whole journey from traditional programming to latest Large Language Models

Although the slide deck is in English, this talk can be presented in Marathi or Hindi as well.