Knowledge that reaches researchers wherever they are.

Web Interface

Full-featured browser access for desktop research.

The browser is the primary surface. The Testaria morphological search engine, the Nomia library, and this site all run in the browser with nothing to install — the natural home for desktop research workflows.

How it works

• Faceted filters over the morphological character matrix narrow the universe of candidate species as you select what you observe.
• Each result is a ranked candidate with full provenance, an exportable identification sheet (CSV), and a reliability report.
• Always the latest data — no client to update, no version drift across a lab.

Mobile App

Field-ready identification on iOS and Android.

The identification filter, carried into the field. On a research vessel, a shoreline transect, or a sampling campaign, the same catalogue travels with you.

How it works

• Run morphological filtering and capture field notes against the same standardized catalogue used on the web.
• Designed for intermittent connectivity — capture in the field, reconcile when you are back online.
• Specimen observations sync to the shared data layer with provenance intact.

Voice Agent

Hands-free identification, spoken at the microscope.

Voice in the microscope. When both hands are on the stage and the specimen is under the objective, you describe what you see out loud instead of breaking focus to type.

How it works

• Speech-to-text (Whisper-class transcription) captures your spoken description of the specimen.
• A natural-language → structured-query bridge translates that description into morphological facet filters.
• You review and correct the interpreted query before it runs — the agent never guesses silently — and candidates can be read back to you.

APIs & MCP

Programmatic access for pipelines and AI agents.

The catalogue as infrastructure. Beyond the human interfaces, the platform exposes its data to software — from research scripts to autonomous AI agents — through open, standards-based endpoints.

How it works

• Query the catalogue from notebooks and pipelines via REST / GraphQL, and expose it to AI agents through the Model Context Protocol (MCP).
• Darwin Core and ColDP exports feed GBIF and OBIS directly; citable DOI releases version every snapshot.
• A polite-pool edge layer with rate-limiting and caching keeps programmatic access fair. The WoRMS MCP is already built; the platform's own API/MCP is in design.