Domain adaptation for Large Language Models

Journey of how we created a domain specific large language model - model selection, data preparation, continue pretraining, fine tuning and RLHF
Who is this presentation for?
Developers and builders working on LLMs
Prerequisite knowledge:
Python, LLM concepts
What you'll learn?
How to build domain specific LLMs

As a seasoned AI technologist and leader with over 20 years of experience, including 10+ years at the forefront of enterprise AI innovation, I specialize in architecting and operationalizing cutting-edge solutions in Generative AI, Agentic AI, Deep Learning, and Large Language Models. My expertise lies in - Driving AI Strategy & Vision:Defining and executing AI roadmaps that deliver transformative business value and competitive advantage, from R&D to full-scale deployment. Building High-Impact AI Systems: Proven success in developing solutions like domain-specific LLMs, AI-powered healthcare copilot. Leading & Scaling AI Teams: Adept at assembling, mentoring, and leading high-performance teams of AI/ML experts, fostering a culture of innovation, research, and excellence. Delivering Measurable Outcomes: Consistently delivered significant results, such as improving document classification accuracy for a platform processing million pages annually, and architecting an AI strategy that improved customer success. I am passionate about translating complex AI research into practical, scalable, and ethical enterprise applications, driving innovation across diverse industries like healthcare, finance, and technology.