AI Concierge Matching

Semantic CV-to-job matching powered by Gemini embeddings and ChromaDB. Extract verified competencies automatically and score candidates against your listings with AI precision.

How it works

Embed CVs

Submit a candidate CV or resume. Our engine generates a dense semantic embedding via Google embedding-001 model.

Match Jobs

Embedding similarity search against your job listings in a tenant-isolated ChromaDB collection. Returns top-5 matches with distance-based scores.

Extract Competencies

Gemini 1.5 Flash reads the CV in context of the job requirements and returns a list of verified competencies present in the candidate.

Pricing

Starter

$199

per month

1,000 CV matches/mo
Competency extraction
Tenant-isolated ChromaDB
REST API access

Growth

$499

per month

10,000 CV matches/mo
Bulk matching endpoint
Custom embedding model
99.9% SLA

Enterprise

Custom

per month

Unlimited matches
Fine-tuned matching model
On-premise ChromaDB
Dedicated support

API example

POST /api/v1/match
x-ledger-api-key: fmr_your_key
Content-Type: application/json

{
  "cvText": "Senior full-stack engineer with 8 years of experience in React, Node.js, and PostgreSQL...",
  "jobId": "job_abc123"
}

→ 200
{
  "score": 0.87,
  "verifiedCompetencies": [
    "React (8 years)",
    "Node.js (8 years)",
    "PostgreSQL (5 years)",
    "System Architecture",
    "Team Leadership"
  ]
}

FAQ

What embedding model do you use?
Google embedding-001 via LangChain, with Gemini 1.5 Flash for competency extraction. Enterprise customers can configure alternative models.

Is my candidate data private?
Yes. Each tenant gets an isolated ChromaDB collection. CV text is used only for embedding generation and is not stored beyond the match request.

How accurate is the matching?
Distance-based scoring produces a 0-1 normalized score. In benchmarks, top-5 recall exceeds 92% for technical roles with detailed CVs.