Executive Recruiter Agent

Autonomous AI recruiting system — sources, screens, engages, and hires candidates with minimal human intervention.

Architecture

┌─────────────────────────────────────────────────────┐
│                   Telegram Bot                       │
│              (Commands & Notifications)              │
└──────────────────────┬──────────────────────────────┘
                       │
┌──────────────────────▼──────────────────────────────┐
│                 Orchestrator                         │
│         (Pipeline State Machine)                     │
├─────────┬──────────┬──────────┬──────────┬──────────┤
│ Sourcer │  Parser  │  Scorer  │ Outreach │  Offers  │
│         │          │          │ + A/B    │          │
├─────────┴──────────┴──────────┴──────────┴──────────┤
│              Scheduling (Google Calendar)             │
├──────────────────────────────────────────────────────┤
│          Analytics & Cost Tracking                    │
├──────────────────────────────────────────────────────┤
│         PostgreSQL / SQLite Database                  │
└──────────────────────────────────────────────────────┘

Features

• Multi-tenant — isolated client data, branding, and credentials
• Candidate sourcing — LinkedIn, Indeed, GitHub, referrals, vault recall
• AI resume parsing — structured profile extraction via Claude
• Intelligent scoring — weighted skill, level, education, and location matching
• Outreach sequences — 4-step drip campaigns with A/B testing
• Pipeline management — enforced state machine with full event history
• Interview scheduling — Google Calendar integration with Meet links
• Offer generation — templated PDF offers with approval routing
• Telegram bot — commands for status, pause, advance, skip, analytics
• Cost tracking — per-candidate/per-hire costs with ROI vs human recruiter
• Security — credential encryption, input sanitization, rate limiting

Quickstart

Prerequisites

• Python 3.11+
• PostgreSQL 15+ (or SQLite for development)
• Anthropic API key
• Telegram bot token

Setup

cd agents/resume
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

export ANTHROPIC_API_KEY="sk-..."
export TELEGRAM_BOT_TOKEN="..."
export DATABASE_URL="postgresql://..."  # or use SQLite default

python3 -m src.db.migrate
python3 -m pytest src/tests/ -v

See docs/CLIENT-ONBOARDING.md for full setup.

Module Overview

Module	Description
src/models/	Client, Candidate, Pipeline data models with CRUD
src/sourcing/	Candidate sourcing and vault recall
src/parser/	Resume parsing (PDF/DOCX + Claude)
src/scoring/	Weighted candidate-job fit scoring
src/outreach/	Email sequences, A/B testing, reply/bounce detection
src/scheduling/	Google Calendar interview scheduling
src/offers/	Offer letter generation and approval routing
src/notifications/	Telegram bot commands and push notifications
src/analytics/	Nightly aggregation, cost tracking, ROI
src/security/	Audit, credential encryption, input validation
src/db/	Connection management, migrations
src/config/	Environment-based settings

Documentation

• Architecture — sub-agents, data flow, schema
• Deployment — setup, env vars, production
• Client Onboarding — adding new clients
• API Reference — commands, programmatic API

Directory Structure

src/           Application source code
docs/          Documentation
marketplace/   Deployment and onboarding tools
templates/     Offer letter templates
archive/       Legacy scripts and data (preserved)
memory/        Agent memory files
research/      Research notes
information/   Reference materials