Evidence-First Hiring

Receipted

Resumes are self-reported. Receipted replaces them with verified evidence — every skill claim backed by real proof, extracted from your stories and scored by AI.

Backend demo coming soon

The Problem

Resumes are vibes. Receipted is proof.

Anyone can write "led cross-functional teams" or "drove 40% growth." Receipted forces the question: where's the receipt? You write the story of what you actually did. The system extracts the claims, verifies them against external sources, and publishes a trust-scored profile recruiters can actually evaluate in under 2 minutes.

Job Seeker Flow

Write stories, not bullet points

You describe what you actually did. Receipted's AI extracts individual claims from your stories, verifies each one against real sources, and converts them into a scored evidence profile.

Verification Engine

Every claim gets a receipt

Claims are verified via You.com's search API and scored on confidence (0–1). Each skill carries a status: verified, unverified, or suspicious — no more self-reported fluff.

Recruiter Flow

Paste a JD, rank candidates instantly

Recruiters paste a job description, the system extracts required skills, ranks all candidates by verified match strength, and drafts personalized outreach emails via Composio Gmail.

Trust Score

Receipt Strength: 0–100

A composite score built from evidence strength (70%), skill diversity bonus (20%), and inflation penalties. Profiles earn a High / Medium / Low trust rating that recruiters can actually rely on.

Architecture

Graph Data Model

PersonStoryClaimSkillProposedSkill

Each Claim node stores verification status, confidence score (0–1), source citations, and a context tag — Operations, Leadership, Risk, Budget, Data, or Customer Service. The graph model makes it trivial to query "which candidates have verified leadership claims with confidence above 0.8."

Tech Stack

What It's Built On

FastAPIPython 3.11Neo4j 5.xOpenAIYou.com APIComposioDocker ComposePydantic

Want to see it run?

Backend runs on FastAPI + Neo4j via Docker Compose. Drop me a line and I'll spin up a live demo.

Get in touch