What speeDB does
speeDB reads a GitHub or GitLab repository, finds every database query across your codebase, and proposes rewrites that execute faster while returning exactly the same results. The challenge with AI suggestions is trust — speeDB solves this by validating every suggestion against the actual repository files, schemas, and line ranges before showing it to you.
Key Features
- Checked Against Real Code — Each cited file path, line range, and quoted schema is re-verified against fetched repository bytes. Suggestions citing non-existent files are automatically dropped and accounted for.
- Strict Same-Output Classification — Optimizations guaranteed to return identical results live in an isolated section with an explicit equivalence proof (covering rows, columns, ordering, NULL semantics, and duplicate handling).
- Comprehensive Engine Coverage — Supports relational databases, distributed SQL, cloud warehouses, document & wide-column stores, key-value caches, graph, vector, and time-series engines under their specific query rules.
- ORM & Call-Chain Analysis — Directly inspects ORM method chains (Prisma, TypeORM, SQLAlchemy, Hibernate, ActiveRecord) to spot N+1 queries, avoidable round trips, and over-fetching that text searches miss.
- Vector Search Recall Safety — Recognizes approximate vector search mechanics and flags any recall-impacting parameter change as a behavioral alteration rather than a same-output tweak.
- Bring Your Own AI (Or None) — Run 100% on-device using Chrome's built-in Gemini Nano model, or connect your own OpenAI, Anthropic, or Google Gemini API key. Zero speeDB servers or tracking.
- Fast Commit Caching & Export — Results cache for 60 minutes per commit for instant revisits, and can be exported as clean Markdown, HTML, or JSON reports.
Who it's for
Backend developers, database administrators, and QA engineers looking to eliminate query bottlenecks, audit pull requests for N+1 performance regressions, and ensure rock-solid data layer efficiency.
FAQ
Is speeDB free? Yes — speeDB is free to install from the Chrome Web Store.
Does speeDB connect to my production database? No. speeDB has no database connection and never executes queries. It analyzes query code and ORM call chains directly from repository files.
Are private repositories safe? Yes. Analysis happens in your browser. With Chrome's built-in AI (Gemini Nano), nothing leaves your machine. With cloud models, requests go directly from your browser to your chosen provider using your own key.