n8n, Supabase
Hiring Agent
Built for Break Into VC - a hiring and training firm simplifying how talent is matched to roles in venture capital and finance.
Automating resume screening with intelligence and precision
This agent evaluates thousands of resumes against each new job description to instantly surface the most relevant candidates - complete with reasoning and scoring. It replaces hours of manual screening with a consistent, data-driven process that recruiters can trust.
Results at a glance
200+ hours saved every month
3.5 minutes average runtime per query
1,200+ resumes processed with accuracy
The problem
The recruitment team at Break Into VC was spending hours manually reviewing resumes for every new role. As the number of candidates grew, the process became increasingly unmanageable, making it difficult to maintain speed and fairness in early-stage screening.
The solution
We built an intelligent agent that automates Level 1 screening for every incoming JD. The system stores all candidate resumes in a centralized database, evaluates each new JD against the entire pool, and returns a ranked shortlist with match scores and explanations. Recruiters simply send a Slack message containing the JD and receive a Google Sheet with the top candidates in minutes.
How it works
Database setup: Resumes uploaded in PDF format are stored in Supabase and indexed using vector embeddings.
Hard match: An SQL query filters candidates by mandatory fields like experience and location.
Relevance scoring: A custom RAG pipeline analyzes semantic similarity between the JD and each candidate profile.
Combination logic: The system merges both filters to produce a final ranked list of 20–50 resumes with scores and reasoning.
Delivery: A Slack command triggers the workflow and shares the resulting Google Sheet automatically.
Impact
Reduced manual screening time by 4 hours per JD
200+ recruiter hours saved monthly across 50 JDs
3.5-minute average end-to-end processing time
Consistent, explainable results that increased recruiter trust in automation
Build overview
Duration: 2 weeks (strategy, setup, database configuration, custom RAG implementation)
Stack: n8n, Supabase, OpenAI, Google Suite, Slack API
Client: Break Into VC (VC hiring and training firm)
Key challenges: Implementing conditional logic for incomplete resumes and optimizing vector retrieval accuracy
In summary
The Hiring Agent transformed resume screening from a manual bottleneck into a fully automated, explainable workflow. By combining structured SQL filters with vector-based reasoning, the team at Break Into VC reduced screening time by over 200 hours a month - while gaining faster, data-backed hiring decisions.
Next work
