We're two founders who got tired of arguing about what a concept means.

The bet

Concepts will be managed like code — but by people who don't write code.

For five years, every team we worked with had the same problem. The glossary lived in a Confluence page nobody updated. The taxonomy lived in a Google Sheet nobody owned. The ontology — if anyone said the word — lived in a Protégé file two senior architects protected like a relic.

Meanwhile every engineer on the team was happily versioning, reviewing, tagging, and reverting code on a daily basis. The same primitives that made software collaboration work since 2008 had simply never made it across the aisle to the people modelling meaning.

Semlify is what happens when you take the muscle memory of GitHub and apply it to a typed graph of concepts — with a UI that a non-engineer can use without ever seeing a `git diff`.

Founders

Two people, complementary backgrounds.

Guiding principles

What we ship and what we don't.

1. 1

   The graph is sacred.

   Every change is auditable, reversible and attributable. We append; we don't destroy.

2. 2

   One engine, any depth.

   Three classes or three hundred — same product, same storage, same API. Users don't grow out of a 'simple mode'; they just model more.

3. 3

   Make the hard thing visible.

   Hierarchy cycles, conflicting definitions, broken references, orphan concepts — the validation panel surfaces them. We don't hide complexity behind a friendly emoji.

4. 4

   Respect both sides of the keyboard.

   Same product must feel natural to a non-technical curator and to a staff data engineer consuming the API. Two audiences, one tool.

5. 5

   Interoperate with standards.

   SKOS, OWL, RDF, JSON-LD. We speak them so customers don't get locked in. The export is faithful enough to round-trip in Protégé.

6. 6

   Boring tech where it doesn't matter.

   React, Postgres, Node.js, Neo4j — well-understood pieces. The magic lives in the model, the versioning, and the UX. Not in stack novelty.