Project Overview
Prism turns the fragmented recruitment process into a cohesive automation layer. It listens to HR inboxes, parses resumes, uses GPT-4 to score candidates, orchestrates interview scheduling via GCal/Gmail, and even drafts final offer/rejection emails based on interviewer feedback. It replaces manual spreadsheet juggling with an intelligent, autonomous pipeline.
System Architecture
Built on n8n as the central orchestrator. Workflows connect Gmail (Trigger/Comms), OpenAI (Reasoning), Airtable (State/Database), and Google Calendar (Scheduling). Webhooks facilitate handoffs between screening, analytics, scheduling, and decision stages.
n8n Orchestrator
Low-code engine managing the 4 core workflows
OpenAI Node
GPT-4 for resume parsing, scoring, and decision drafting
Airtable
Structured database for candidate state and analytics
Google Workspace
Gmail and Calendar for seamless communication
Implementation Details
Code Example
// n8n Code Node Example: Calculate Match Score\nconst candidate = items[0].json;\nconst requirements = ['React', 'Node', 'AI'];\nlet score = 0;\n\nrequirements.forEach(req => {\n if (candidate.skills.toLowerCase().includes(req.toLowerCase())) {\n score += 10;\n }\n});\n\nreturn { json: { ...candidate, match_score: score } };Agent Memory
Always enforce a JSON schema in your System Prompt when asking the LLM to parse resumes. This ensures n8n downstream nodes can reliably map fields like 'years_experience' without regex hacks.
Workflow
Intake: Email triggers workflow; PDF parses to JSON.\n2. Scoring: GPT-4 evaluates fit; updates Airtable.\n3. Analytics: Webhook triggers deep-dive skill matching.\n4. Schedule: Candidate selects GCal slot; invite sent.\n5. Decision: Feedback aggregated; AI recommends Hire/No-Hire.
Results & Impact
"Prism replaced a patchwork of spreadsheets and inbox digging with one coherent AI pipeline. We now spend time talking to people, not chasing info."
Time Saved
Eliminated 15-30m of manual work per candidate.
Fairness
Standardized AI scoring criteria for every applicant.
Velocity
Zero latency handoffs between screening, scheduling, and offers.
About the Author
Akshaay
AI Context Engineer
Apex Neural
Akshaay is an AI Context Engineer experienced in designing agentic AI systems, LLM-powered applications, and automation workflows. He combines modern backend engineering with no-code and low-code automation approaches to deliver production-ready AI solutions deployed on scalable infrastructure.
Contributors
ParmeetReady to Build Your AI Solution?
Get a free consultation and see how we can help transform your business.