From mathematics to robotics, ML technologies are rapidly advancing and reshaping traditional approaches to capturing intelligence. While these techniques are powerful, they are often ad-hoc, opaque and hard to analyse. At Basis, our goal is to understand intelligence and build systems that can reason and learn, and we believe that programming languages technologies, such as program synthesis, probabilistic programming, and algebraic effects can serve as a key building block for principled approaches to reasoning about artificial intelligence. In this talk, we discuss our ongoing research and how we are using programming languages to develop a deeper understanding of intelligence.

Basis is the industry sponsor of the LAFI 2026 workshop.

Kiran Gopinathan is a Research Scientist at Basis. Her research focuses on techniques for developing newer and better tools for automated formal verification – the art of using computers to automagically construct mathematical proofs about the correctness of software. Her research interests cover formal verification, program synthesis, type systems, language design and proof engineering. She previously completed her postdoc with Talia Ringer at UIUC, and before that earned her PhD in Programming Languages Research from the National University of Singapore on automating the maintenance of formally verified software.

bsky: @kirancodes.me