A lightning-fast, local-first knowledge indexer.
Hqllr transforms raw documents into compact, deduplicated knowledge ready for instant search and verifiable answers.
- Local & Sovereign — runs offline; your data stays yours.
- Atomised & Deduped — stores only what’s unique; saves space and time.
- Instantly Searchable — built for rapid lookups at any scale.
- Provenance-Ready — designed so downstream answers can cite and verify.
Preview – architecture and APIs here are high-level; core internals are private while we harden the engine.
- Public demos, schemas, and examples will land first.
- Full ingest features will appear in staged releases.
- Examples:
examples/show how to prepare data and query mock indices. - CLI (preview):
hqllr --help(when published)
- Preview demos & mock datasets
- Public CLI (read-only)
- SDK bindings (Dart/Go/Python)
- Private ingest beta (early access)
- Built for offline/sovereign use. No telemetry by default.
- See
SECURITY.mdandPRIVACY.md.