CostSleuth

CostSleuth is a web app (with optional CLI agent) that monitors cloud data warehouse usage and translates it into actionable, team-level cost controls. It connects to Snowflake, BigQuery, and Databricks SQL to ingest query history, warehouse/slot usage, and job metadata, then attributes spend to owners via tags, dbt models, service accounts, and BI dashboards. The product flags waste patterns (runaway dashboards, unbounded scans, duplicate transformations, over-provisioned warehouses) and recommends concrete fixes like clustering/partitioning changes, materialization tweaks, caching, or schedule adjustments. It also enforces guardrails: budgets, alerts, and “soft limits” that require justification for expensive jobs. This is a combination traditional + AI app: AI helps summarize root causes and propose optimizations, but the core value is accurate attribution and controls. Realistically, you’ll win by being brutally reliable, not “magical.”