Four steps run automatically the moment you drop a file in. No tagging, no foldering, no naming conventions.
PDFs, images and text are parsed locally. Real text is pulled from the document — not guessed from the filename.
An LLM assigns one of six categories with a confidence score, plus structured entities: skills, orgs, roles, dates.
Relationship triples link certification → skill → project → internship → career path into a live force-directed graph.
Hybrid search: metadata filters for exact asks, vector similarity for meaning. Originals always kept intact.
Every module reads from the same knowledge store, so nothing ever goes out of sync.
KPI strip, category distribution and growth trend at a glance.
Interactive D3 force graph of every entity and the edges between them.
Chronological story of your growth, year by year.
Weighted skill profile computed from evidence across all documents.
Ask questions in plain language, answered from your own documents with citations.
A shareable identity page generated from everything above.
That's the whole point. Load the demo identity and try it in ten seconds.