pillar 01 · data + gym
Close the gaps in your article data.
Ingest a read-only snapshot of your EPLAN article database. An LLM closes text gaps — translations, descriptions — against deterministic validators. Every change is reviewed like a pull request; a local client writes approved values back into EPLAN.
- the model proposes, the validator decides
- overlay only — ingested data is never mutated
- approved changes flow back through a local client