Evident: Resume Claims You Can Prove From Code
Mayank Tiwari
Evidence graph: repositories, commits and source files mapped into verifiable technical provenance.
Provenance inspector: tap "Why this claim?" to see the repo, file path, commit hash and code snippet behind a resume bullet.
Application studio: paste a job post and get resume bullets grounded only in evidence, with zero hallucinated claims.
Extracted source evidence: each skill labelled Direct, Supported, Partial, Gap or User-declared, with authorship and freshness.
Defense arena: interview questions generated from your own code, so you can defend every claim you make.
Built for the RevenueCat Shipathon 2026 (Devpost, Next Gen student track).
Students don't lack work. They lack evidence. Their skills are buried in branches and commits, so resumes fill up with AI buzzwords nobody can verify. Evident reads what you actually built and only lets you claim what it can back up.
5-dimensional evidence model. Every skill is tracked by evidence status (direct, supported, partial, gap or user-declared), source type (source file, commit history, dependency manifest, API route), authorship, confidence and freshness.
Honest job matching. Paste a job post and get an evidence coverage matrix ("5 direct, 2 supported, 1 gap") instead of a made-up "87% match".
Every claim is clickable proof. Tap "Why this claim?" to see the exact file path, commit SHA and code snippet behind a resume bullet.
Anti-hallucination auditor. Unsupported claims (say, Kubernetes or AWS with no evidence) are flagged and rejected, and a golden test suite enforces it.
React Native · TypeScript · Zustand · RevenueCat (Pro tier: a living career memory that re-indexes your repos as you push).
Like this project
Posted Oct 16, 2026
Reads your repos and commits, extracts skills with a 5-dimension evidence model, and links every resume claim to the file and commit that proves it.
Likes
1
Views
2
Timeline
Jan 1, 2026 - Dec 31, 2026
Clients
RevenueCat
