RecruiterClaw

Agentic recruiting automation — source → enrich → outreach → schedule

Public overview — source code is private. This repo describes what the product does and how it's built.

---

The problem

Recruiters lose enormous time stitching together sourcing, enrichment, outreach and scheduling across a dozen disconnected tools. Existing automation tools are either brittle, risky to a recruiter's accounts, or can't be customised to a desk's real workflow.

What RecruiterClaw does

RecruiterClaw is an agentic copilot for recruiters that runs the full top-of-funnel as a chainable set of skills:

• End-to-end pipeline — source candidates, enrich profiles, personalise outreach, and book meetings, driven from chat.
• Per-user, authentic execution — automation runs in each recruiter's own browser session and IP, so activity is genuine to their account rather than routed through shared infrastructure — safer and more compliant by design.
• Chainable "skills" — composable building blocks recruiters can combine into custom sequences.
• Hybrid AI — a mix of local and cloud models to balance cost, speed and capability.
• Multi-seat ready — built for teams, with a branded desktop experience.

Status: In active development.

Architecture (high level)

Chat / workflow (the "brain" + queue)
        │  orchestrates skills, never touches third-party sites directly
        ▼
Per-user local agent (recruiter's own browser session & IP)
        │
   source · enrich · outreach · schedule   ─►   messaging integrations

Tech

TypeScript · Node.js · agent runtime (OpenClaw-based) · hybrid local + cloud LLMs (Google Gemini + a local model) · per-user browser-automation architecture · messaging integrations · cloud VM provisioning · Electron / React desktop app · multi-seat billing.

---

Built by Dominic Gonsalves · open to contract / fractional AI roles