Leeroo open-sources Kapso, an engine for iterative enterprise data experiments
Leeroo, founded by Alireza and Majid, has released Kapso as open source on GitHub. The tool runs experiments against enterprise data systems, deploys the best-performing solution, and retains learnings from each run to inform future work. The founders describe using it on projects such as migrating a company with over 30 source-system families and thousands of tables onto a lakehouse.
GoKawiil's interpretation of the reporting above, not reported fact.
The founders argue that understanding large, messy enterprise data estates can't be solved by upfront documentation or static retrieval, but requires repeated hypothesis-testing similar in principle to reinforcement learning. By open-sourcing the engine, Leeroo could be aiming to establish Kapso as a standard approach for AI-driven data engineering, though its real-world adoption and effectiveness beyond the company's own case studies remain unproven.
- Kapso is now open source on GitHub, built by Leeroo's founders.
- It automates iterative experimentation and deployment for enterprise data projects.
- The approach was tested on a migration involving 30+ source systems and thousands of tables.
Source: news.ycombinator.com, 2026-11-12
Published there as: “Show HN: Kapso – Iterative learning from enterprise data”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.