Back to Projects

Roomco

Social and chatting application for young people with AI chatbot integration and content moderation.

View Project
Roomco

Project Overview

Real-time social networking and messaging platform designed for young users. Features WebSocket-based chat, AI-powered chatbots using ChatGPT API, multi-language support with image generation capabilities, and content safety through AI-powered nudity detection.

Role

Lead Developer and Migration Architect

Technologies Used

NestJSTypeScriptSocket.ioMongoDBChatGPT APIFastAPIfast.aiRedis Pub/Sub

Tasks

  • Led complete migration from Express.js to NestJS with TypeScript
  • Architected and supervised migration strategy across development team
  • Participated in server migration to more efficient infrastructure
  • Upgraded Node.js version from 12 to 22 with comprehensive testing
  • Created chatbots using ChatGPT SSE API with streaming responses
  • Managed multi-language bot responses using JSON schema validation
  • Integrated ChatGPT API for image generation in chatbots
  • Developed nudity detection service using fast.ai deployed on external server
  • Implemented FastAPI microservice for real-time image moderation
  • Established Redis Pub/Sub communication between NestJS and FastAPI services
  • Created comprehensive migration documentation and training materials

Achievements

  • Successfully migrated entire codebase from Express.js to NestJS with zero downtime
  • Implemented type-safe architecture reducing runtime errors by 75%
  • Upgraded Node.js across 10 major versions without breaking changes
  • Improved server performance and code maintainability through modern architecture
  • Deployed AI-powered content moderation system protecting young users
  • Achieved real-time image scanning with <500ms response time using Redis Pub/Sub
  • Implemented AI chatbot increasing user engagement by 40%
  • Created scalable microservices architecture supporting independent scaling

Contributions

  • Architected migration plan from Express.js to NestJS with TypeScript
  • Supervised development team through migration process with regular code reviews
  • Designed and documented microservices architecture for content moderation
  • Built fast.ai-based nudity detection model with 95%+ accuracy
  • Implemented FastAPI service for high-performance image processing
  • Established Redis Pub/Sub messaging pattern for inter-service communication
  • Created comprehensive migration guide documenting patterns and best practices
  • Provided mentorship and training on NestJS, TypeScript, and microservices
  • Implemented streaming responses for real-time AI interactions
  • Developed multi-language support system for chatbot responses with type safety