Real-time support chat with AI drafting, routing rules, and Redis-backed message speed.
Designed the messaging layer for low-latency delivery, presence (online/typing), and consistent ordering across devices.
Added AI response drafting that summarizes customer context and proposes replies (agent remains in control before sending).
Built an internal admin area for agent seats, canned responses, routing rules, and conversation tags.
Framework
Language
AI Service
Real-time Communication
Cache & Pub/Sub
Database
Deployment
Socket-based messaging with delivery status, typing indicators, and reconnect resilience.
AI-generated drafts and summaries that help agents respond faster without losing tone control.
Redis-backed caching for recent threads so inbox scrolling and search feels instant.
Auto-assign conversations by tags, keywords, working hours, and agent capacity.
Agent-only notes per thread to keep context without leaking info to customers.
Session-aware sync so agents can move between devices without losing thread state.
Have an idea you want to bring to life? Let's shape it into something you're proud to show.