Partnership opportunities

Save $100 on your pass

Call to action
Your text goes here. Insert your content, thoughts, or information in this space.
Button

Back to speakers

Paras
Chaudhary
Software Engineering Lead, Generative AI, Document Processing & Drafting
Morgan & Morgan
Paras is a skilled software development engineer specializing in Artificial Intelligence with a penchant for entrepreneurship and a passion for utilizing data to drive business growth. He is a proud recipient of the O-1A visa, a category reserved by the U.S. government for individuals at the pinnacle of their professions. Paras holds a Master of Science in Computer Science degree with a specialization in Machine Learning from Columbia University, where he transferred from Carnegie Mellon University. Paras' expertise in applied Generative AI and Machine Learning has been honed through his experiences at Morgan & Morgan (world's biggest injury law firm) and Amazon, both of which required him to convert hundreds of terabytes of unstructured data to actionable structured data. At Morgan & Morgan, Paras is responsible for researching, developing and deploying the company's AI strategy. He is tasked with building Large Language Model based tools that ingest hundreds of millions of documents to reliably provide value to this legal data behemoth; where efficiency is not merely a strategic advantage; it is the linchpin of the company's operational success, ensuring the prudent allocation of resources to benefit the clients and in turn the business stakeholders.
05 June 2025 12:30 - 12:50
Panel discussion: Building a fine-tuning pipeline for LLM alignment
Aligning large language models with specific objectives and use cases requires a robust fine-tuning pipeline that ensures both precision and efficiency. Panelists will discuss the technical intricacies of building such pipelines, sharing insights into data curation, model adjustment strategies, and monitoring alignment, while addressing challenges like maintaining model generalization and mitigating risks associated with bias and drift.