Instacart built Maple, a backend brain for handling millions of LLM prompts—fast, cheap, and shared across teams.

It’s not just another service. Maple runs on Temporal, PyArrow, and S3, strip-mines away provider-specific boilerplate, auto-batches prompts, retries failures, and slashes LLM costs by up to 50%.

Bigger picture: It’s part of a broader shift—companies rolling their own LLM platforms to centralize orchestration and cut cloud costs. Everyone wants ChatGPT, but no one wants to pay retail.