Northeast Scala Symposium
2023

Sisir Koppaka

Sisir Koppaka is a practising Data Scientist and Software Engineer working at the intersection of data and Scala. I've been building AI systems from the ground up for over 10 years, and the role of Scala has only ever expanded in this time frame. I'm especially interested in the recent advances in LLMs and how these can help unlock a variety of new use cases when supported by the Scala language.

The Power of Scala with Large Language Models

Abstract - Scala is an incredibly powerful, type safe language. Large Language Models (LLMs) are the starting point of AI applications today. Integrating Scala with LLMs provides a powerful toolkit that can abstract over a larger variety of data and still retain the benefits of type safety.

The talk provides an introductory overview of Large Language Models (LLMs) and how to use them with Scala. By examining the summarisation capabilities of LLMs, we will show how Scala's features such as pattern matching and type safety can provide adequate safeguards and expansion of scope for accommodating new kinds of use cases given different kinds of data. LLMs can be seen as "extending the compiler" to provide new input data the same type safety capabilities given to Scala code.