AI Resume Shortener & Recruitment Intelligence Platform

AI Resume Shortener is an enterprise-grade recruitment intelligence and resume optimization app for Frappe HR / ERPNext. Automate candidate resume screening, ATS scoring, and interview question generation for Frappe HR with multi-model AI failover and token cost tracking across Google Gemini, OpenRouter, OpenAI, and Claude.

📖 Full Documentation Portal: Explore the interactive Product Overview, read the detailed User Guide, or check out the Product Architecture Overview.

Key Features

1. Multi-Model AI Orchestration & Failover

• Dynamic AI Models: Configure and prioritize multiple AI models in the Ai Model Config List child table without hardcoded model names.

• Multi-Provider Support: Built-in support for Google Gemini (gemini-2.5-flash, gemini-2.0-flash) and OpenRouter (gpt-4o-mini, claude-3-haiku, etc.).

• Automatic Failover: If the primary AI provider fails or times out, the system automatically falls back to secondary enabled providers.

• Cost & Token Tracking: Calculates exact input/output token usage and estimated USD cost per resume evaluation.

2. Candidate Evaluation & ATS Matching

• Job Description Alignment: Matches resume contents against standard Job Description requirements.

• Structured Metrics:

  • Overall Fit Score (0–100)

  • Match Percentage (0–100%)

  • Confidence Score (0–100%)

  • Category breakdown: Education, Experience, and Skills scores

• Candidate Star Rating: Automatically sets Frappe HR candidate rating (1 to 5 stars) based on recommendation level and overall fit score.

• Generated Insights: Extracted skills, missing skills, candidate strengths, summary, and auto-generated interview questions (Technical, HR, and Behavioral).

3. Enterprise Prompt Management & Versioning

• Ai Prompt Version DocType: Manage versioned prompt templates (v1.0-default, v2.0-custom) with one-click activation and rollback.

• Audit Traceability: Every evaluation logs the exact prompt version, AI model, and provider used.

4. Background Processing & Queue Locks

• Asynchronous Execution: Per-applicant background jobs enqueued via frappe.enqueue() on queue "default".

• Non-blocking Hourly Scheduler: Automatically picks up unanalyzed applicants without clogging server resources.

• Queue Lock Protection: Built-in deduplication safeguards prevent accidental double-enqueuing or duplicate API executions.

5. Auditability & Analysis History

• Ai Analysis Log: Centralized audit repository recording job status (Queued, Success, Failed, Retried), execution time (seconds), token counts, costs, error logs, and raw JSON payloads.

• Job Applicant AI History: Embedded child table on Job Applicant keeping a complete record of historical evaluations over time.

• SHA256 Resume Hash: Detects duplicate CV submissions across applicants.

6. Modern Desk Workspace

• Intelligence Center Workspace: Interactive workspace featuring Number Cards (Total Applicants, Analyzed, Pending, Total Logs), Shortcuts, and Card Breaks.

• Instant Desk Actions:

  • Analyze with AI: Instant synchronous evaluation with UI progress overlay.

  • Enqueue Background Job: Asynchronous background worker enqueuing.

  • Benchmark Models: Runs side-by-side comparative evaluations across all enabled AI models.

Configuration

1. Navigate to AI Resume Shortener ➔ AI Configuration.

2. Open Ai Provider and set your API keys:

   • Google Gemini: Set API Key for Google AI Studio.

   • OpenRouter: Set API Key for OpenRouter API.

3. Open Ai Model Config List to enable your preferred models and set token pricing (per 1M tokens).

4. Select the default AI model in Ai Config.

Verification & Usage

1. Open any Job Applicant document with a resume attachment (.pdf, .docx, .txt) or link.

2. Click Analyze with AI for instant evaluation, or select Actions ➔ Enqueue Background Job for background processing.

3. Review candidate scores, ratings, missing skills, generated interview questions, and historical evaluations.

4. Check Ai Analysis Log to monitor execution times, token consumption, and API costs.