Autoblocks gives AI product teams a workspace for testing, evaluating, and monitoring LLM-powered applications before and after launch, aimed at catching failures that a single manual review would miss. The company was founded in 2023 by Adam Nolte, Haroon Choudery, and Hamza Choudery, and has raised $2 million from The House Fund and Kyber Knight Capital.
The platform runs automated tests across thousands of scenarios and edge cases, capturing realistic user interactions rather than relying on a fixed set of manually written test cases, and integrates subject-matter-expert (SME) feedback directly into the evaluation loop. Autoblocks is built with HIPAA and SOC 2 Type 2 compliance, positioning it for regulated industries like healthcare and finance where an AI mistake carries higher stakes.
Product and engineering teams shipping customer-facing AI chatbots and agents, particularly in regulated sectors, are the primary users, more than solo developers testing a hobby project. Pricing starts with a free tier for building and testing, moves to a contact-priced Pro plan with full evaluation and prompt-management features, and an Enterprise plan with custom terms based on data volume and evaluation cycles.
