Platform Overview
AIQLick is an AI-powered recruitment platform that connects employers and job seekers through intelligent talent management, automated CV analysis, skill-based job matching, and integrated video interviewing. It supports multi-company collaboration, pipeline tracking, and a full billing system with credit-based AI operations.
Key Capabilities
- Talent Management -- Store, organize, and search candidate profiles across talent pools with fine-grained access control and external sharing.
- AI-Powered CV Parsing -- Upload resumes and extract structured data (skills, experience, education) using AWS Bedrock LLMs with real-time progress streaming.
- Job Matching -- Score candidates against job requirements using hybrid matching (vector similarity + full-text search) with configurable weighting.
- Video Interviews -- Conduct live interviews through a custom Jitsi Meet integration with real-time transcription and AI-generated meeting insights.
- Pipeline Management -- Track candidates through configurable Kanban-style workflows from application to hire.
- Multi-Company Collaboration -- Users belong to multiple organizations with role-based permissions. Share talent pools, pipelines, and candidate lists across teams.
- AI Agent -- Conversational assistant with RAG-powered document search, tool execution, and streaming responses.
- Credit-Based Billing -- Stripe-integrated subscription and credit system where 1 credit equals $1 of AI token consumption. Enterprise customers can use a post-paid model with monthly usage-based invoicing.
Architecture Summary
AIQLick runs as four interconnected services that share a single PostgreSQL database per environment:
| Service | Role | Port |
|---|---|---|
| Backend | Core API server handling authentication, business logic, billing, and notifications | 4001 |
| Frontend | Web application providing the user interface for employers, job seekers, and administrators | 4000 |
| Background Tasks | AI/ML pipelines for CV extraction, job matching, transcription processing, and agent chat | 8000 |
| Jitsi Meet | Video conferencing with real-time transcription via a custom Jigasi gateway | -- |
The backend and background-tasks services share the same PostgreSQL database. Schema is managed exclusively through Prisma in the backend; background-tasks connects with raw AsyncPG queries. Redis provides queue infrastructure (Bull) and pub/sub for real-time subscriptions.
Tech Stack
| Layer | Technologies |
|---|---|
| Backend | NestJS 11, Apollo Server 4 (GraphQL), Prisma 6, PostgreSQL 16, Redis, Stripe, Bull queues |
| Frontend | Next.js 16 (App Router, Turbopack), Apollo Client, custom TW* components, Tailwind CSS 4, Zustand |
| Background Tasks | FastAPI, Strawberry GraphQL, AsyncPG, AWS Bedrock (Claude), LangChain, pgvector |
| Video | Custom Jitsi Meet fork (React/Webpack), custom Jigasi fork (Java) for Whisper transcription |
| Infrastructure | AWS EC2, RDS, S3, CloudFront, ECR, Amplify, Bedrock, Rekognition, Transcribe |
Production URLs
| Service | Production | Development |
|---|---|---|
| Frontend | https://www.aiqlick.com | https://dev.aiqlick.com |
| Backend API | https://api.aiqlick.com | https://api-dev.aiqlick.com |
| AI Services | https://ai.aiqlick.com | https://ai-dev.aiqlick.com |
| Video Meetings | https://book.aiqlick.com | -- |
| File Storage | https://storage.aiqlick.com | https://storage-dev.aiqlick.com |
All services are deployed on AWS in the eu-north-1 (Stockholm) region. The frontend is hosted on AWS Amplify with automatic deployments on push. Backend services run as Docker containers on EC2 instances, with images stored in ECR.