unLAP offers an AI‑driven pipeline that monitors BigQuery workloads, automatically rewrites expensive OLAP queries, and validates the optimizations with formal proofs and regression tests. The verified rewrites can be deployed with a single click or automatically, hot‑swapping queries without code changes to reduce cloud costs and maintain development speed.
Funding
Funding not disclosed
Founders
Product
Problem
Expensive OLAP queries that slip into production cause high cloud costs and require manual tuning, which slows development cycles and creates cross‑team friction.
Solution
unLAP provides an autonomous, AI‑driven pipeline that continuously monitors data warehouse workloads, identifies costly queries, and generates optimized SQL rewrites. Each candidate rewrite is backed by formal proofs and regression tests in a managed sandbox to ensure functional correctness and performance gains. Verified optimizations are presented to engineers for a one‑click approval or can be auto‑deployed, hot‑swapping the original query without any code changes. By integrating via a single import statement, the system works with existing BigQuery environments and eliminates the need for manual query triage, reducing cloud spend while preserving development velocity.
Target Audience
Primary users are platform engineering teams, data platform owners, and FinOps groups managing large‑scale OLAP warehouses who need to control query costs without disrupting application development.
Features
- Continuous observation of query workloads in BigQuery to detect high‑cost, templated queries
- AI generation of optimized SQL candidates that adapt to data layout and apply techniques such as predicate pushdown, join elimination, and window‑function rewrites
- Formal verification layer with mathematical proofs and automated regression tests against the original query results
- One‑click or automated deployment that hot‑swaps the optimized query in production with zero code modifications
- Drop‑in SDK requiring only a single import change to activate the full optimization pipeline
- Detailed cost and performance impact reporting to quantify savings before deployment