Anirudh Shrinivason (Cohere) – LLMs for Out-of-Domain Use Cases

This talk will “delve” into Cohere’s strategies for harnessing the potential of Large Language Models (LLMs) within the enterprise domain. By exploring fine-tuning techniques and the integration of Retrieval Augmented Generation (RAG), we aim to showcase how these methods can be tailored to solve complex, industry-specific challenges. The presentation will cover the considerations to be taken when trying to solve complex enterprise grade out-of-domain language model use cases and introduce some concepts that Cohere’s applied ML teams make use of to solve out-of-domain problems.

This seminar is open for students who are taking 50.045 Information Retrieval course only.

Anirudh Shrinivason
Software/AI Engineer
Cohere

Anirudh Shrinivason is a Software/AI engineer with a technical background in data platform development and applied ML. Holding a Bachelor of Engineering in Computer Science and Design from the Singapore University of Technology and Design, he started off as a data engineer in Grab, and now works at Cohere as a member of technical staff. Anirudh has worked on various projects such as building robust data platforms, enhancing data discovery and complex genAI solutions in the data domain. He has also led ML initiatives, working on projects such as recommendation engines, multi-agent setups and developing on genAI frameworks. Anirudh is passionate about greenfield topics and is currently interested in making use of LLMs to solve complex use cases and hard problems.