Kiln

Conversationally measures real skill proficiency against any job description, then prescribes the shortest path to close the gap.

Scroll to begin

How Kiln Works

Five agents, one verdict.

A deterministic supervisor routes between specialized agents. The intelligence lives in the agents; the supervisor is a 30-line if/else.

Parser

Deterministic dictionary scan extracts skills from JD + résumé. Framework evidence inherits to underlying languages (LangChain → Python ≥ 3).

Strategist

Picks which skills are worth probing for the role, weighting by JD importance and gap size. The first question is targeted, not generic.

Interviewer ↔ Evaluator

Asks adaptive follow-ups, scores each rubric dimension with quoted evidence. Probes deeper until confidence ≥ 0.7 or the question budget is hit.

Architect

Ranks gaps by value-per-hour-of-study and grounds every recommended resource via Tavily + Fetch — no hallucinated course names, no broken links.

See what someone actually knows.

Paste a job description, drop in a résumé, and let Kiln run the assessment in your browser. Every score comes with quoted evidence and a calibrated confidence value.