Hazel - Paris Vibe-a-thon: “Build a One-Person Unicorn” — November 15, 2025
AI Tinkerers - Paris
Hackathon Showcase

Hazel

Building an AI agent that will replace HR agency from sourcing to interview. We’re a team of 2, we need an extra AI engineer.

2 members

Hazel is an end-to-end AI recruitment agency that automates the entire hiring pipeline for startups and companies without HR teams. It solves the core problem of traditional recruitment: agencies are slow, expensive, inconsistent, and human-dependent. Hazel replaces them with coordinated autonomous AI agents that handle everything from job creation to interview scheduling.

Our target customers are startups, scaling teams, and companies without dedicated recruiters who need fast, affordable, and high-quality hiring.

Hazel’s main features include a Job Offer Agent that writes and structures job descriptions, a Sourcing Agent that searches and matches talent across multiple sources, an Outreach Agent that contacts qualified candidates and manages replies, and a Scheduling Agent that books interviews autonomously. All agents communicate through n8n workflows, retry automatically, and operate continuously. The user interacts through a simple natural-language interface that allows them to refine requirements and oversee progress.

Demo Flow:
The demo shows (1) uploading or describing a job role, (2) Hazel generating the job offer and search criteria, (3) the sourcing agent finding and ranking candidates, (4) automated outreach messages being sent, (5) candidates responding, and (6) the scheduling agent booking interviews and producing a final shortlist. This demonstrates real execution, agent autonomy, and a complete hiring loop.

Execution & Functionality:
Hazel is fully functional as an automated pipeline. Each agent runs independently, communicates with others, handles failure modes (e.g., restarting a search), and escalates outputs to the UI layer.

AI & Agents Usage:
We implemented multiple agents using Mistral AI for reasoning, n8n for orchestration, and internal state-sharing logic. The system retries failed searches, regenerates job specs, and continues until candidates are identified—fully autonomously.

Scalability (One-Person Model):
Hazel is built to scale infinitely with almost zero marginal cost per hire. A single person can manage and improve the system because the agents handle the full operational workload.

Problem Clarity & Market Impact:
The recruitment industry is a ~$900B global market dominated by legacy agencies. Hazel delivers 8–10× lower cost, faster hiring, unbiased screening, and 24/7 operation. This model directly replaces agencies and has massive market impact.

Product Narrative:
The story is simple: upload a role → watch four agents operate autonomously → receive a curated shortlist → hire. Hazel turns hiring from a manual process into an automated service.

Partner Tech Usage:
We used Mistral AI for reasoning, n8n for agent orchestration, Lovable for generating the front-page design, Figma for visuals, and Cursor for code development.

AI Tinkerers ElevenLabs Figma Google Lovable Mistral AI n8n