Boundary AI

Boundary AI is a comprehensive toolkit designed to streamline tasks for AI engineers. It leverages a special configuration language known as BAML (Basically, A Made-up Language) to improve the performance of Large Language Models (LLMs).

With BAML, AI engineers can transform complex prompt templates into typed functions that are simpler to execute and test, removing parsing overhead and type-related errors.

Using an LLM with BAML is similar to calling a standard function. Boundary AI also facilitates rapid testing of new prompts within various IDEs, including BAML's VSCode Playground UI.

Additionally, the toolkit features Boundary Studio, which monitors and tracks the performance of each LLM function over time. BAML is primarily coded in Rust and supports Openai, Anthropic, Gemini, Mistral, and user-provided models, with plans to incorporate non-generative models.

BAML deployment generates Python or Typescript code. Unlike other data modeling libraries, BAML is uniquely typesafe and always keeps prompts visible. It includes a built-in playground and supports any model.

The BAML compiler and VSCode extension are free and open source, with paid services available for the monitoring and optimization features of Boundary Studio.