Teamstar is a multi-audience social networking platform built for professionals, educators, students, healthcare providers, and athletes. The product sits in the same category as LinkedIn and Facebook but targets verticals those platforms underserve — coaching networks, healthcare collaboration, academic communities, and sport organisations.
I worked on backend architecture and feature delivery across the platform, covering authentication, real-time messaging, the feed engine, business profiles, networking workflows, and the GraphQL/REST API layer that powers both the Quasar web app and the Flutter mobile clients. The system runs on AWS with Laravel Horizon handling queue throughput, WebSockets for real-time events, and Redis for caching and presence tracking.
The scale challenges were the interesting part: feed fan-out with privacy tiers, presence and read-receipt accuracy across web and mobile sessions, and a global search layer covering people, hashtags, events, groups, forms, and feeds without dragging query latency. AWS Rekognition handles image moderation on uploaded media, and Generative AI / ChatGPT integration powers content assistance inside the composer.
Key Features
- Authentication — password login with OTP verification, session management across web and mobile.
- Profiles — personal profile customisation plus tiered Business Profiles with three admin levels (full edit, organisational data only, content/social only).
- Global search — unified query across people, feeds, hashtags, events, groups, and forms.
- Connections & follows — dual graph model (mutual connections plus asymmetric follows) similar to LinkedIn + Twitter.
- Feed system — Facebook/LinkedIn-style feed with privacy controls, hashtag indexing, and real-time updates via Laravel WebSockets.
- Direct & group chat — WhatsApp-style messaging with read receipts, online/last-active tracking, and media attachments.
- Audio & video calls — VideoSDK integration for one-to-one and group calls.
- Notifications — real-time in-app notifications plus push delivery via Firebase.
- Networking services — structured request flows for mentoring, outreach, partnerships, and opinion exchange.
- Events & calendar — scheduling system with iCal export and event-level privacy controls.
Tools & Technologies
Backend: Laravel 12, PHP 8.2, Node.js, Express.js, GraphQL, REST APIs, Laravel Horizon, Queues, WebSockets
Frontend (Web): Flutter Web
Admin Frontend (Web): Vue.js, Quasar Framework
Mobile: Flutter
Data: MySQL, Redis
Infrastructure: AWS (EC2, S3, CloudFront), Nginx migrated to Digital Ocean
Real-time & Media: Laravel WebSockets, VideoSDK, Firebase (push notifications)
AI / ML: AWS Rekognition (image moderation), ChatGPT / Generative AI (content assistance)
Project Gallery
Screenshots and visuals from the project