Smart Routing: Seamless Conversation Management The Problem Small businesses in India often juggle customer conversations acrosspero WhatsApp Business, Instagram, SMS, Web Chat, and Email. Each channel lives in a separate app or inbox, f...
Smart Routing: Seamless Conversation Management
The Problem
Small businesses in India often juggle customer conversations acrosspero WhatsApp Business, Instagram, SMS, Web Chat, and Email. Each channel lives in a separate app or inbox, forcing agents to switch contexts constantly. This fragmentation slows down response times, increases the chance of missing a message, and makes it hard for owners to see a single view of their customer interactions.
What Changes for the Business
With GoSumo’s Smart Routing feature, all incoming messages are collected through a single AI‑driven interface. Agents no longer need to open multiple apps; they see every conversation in one placeリー. The system automatically classifies the intent of each message and forwards it to the correct channel or the most appropriate agent, ensuring that customers receive timely, consistent replies.
How It Works
1. Unified Ingestion
GoSumo’s NestJS backend receives messages from the integrated channels via webhooks. Every payload is normalized into a common format and stored in PostgreSQL through Prisma. Redis caches recent conversations to reduce database load.
2. AI‑Powered Intent Detection
OpenRouter AI processes the conversation text, generating an intent vector. The vector is stored in Qdrant, a vector database, alongside the conversation metadata. This allows the خور system to perform similarity searches and identify the best‑matched response templates or live agents.
3. Smart Routing Logic
The routing engine, written in TypeScript, evaluates the intent vector against predefined rules. If the intent matches a high‑confidence automated response, the message is forwarded to the appropriate channel with the prepared reply. If the intent requires human intervention, the engine assigns the conversation to an agent based on availability, skill set, and workload. BullMQ queues the routing task, ensuring reliable delivery even under high traffic.
4. Multi‑Tenant Support
Each Indian SMB operates in its own tenant namespace. Tenant data isolation is enforced at the PostgreSQL schema level, while Redis keys are prefixed with the tenant ID. The routing logic respects tenant boundaries, ensuring that conversations from one business never leak into another.
5. Real‑Time Dashboard
A Next.js front‑end displays all conversations in a single pane. Agents can view the conversation history, see the AI’s suggested routing, and override it if needed. The dashboard also shows real‑time metrics such as average response time and agent workload, giving owners a clear view of their client management performance.
Using Smart Routing
Set Up Channels – Connect WhatsApp Business, Instagram, SMS, Web Chat, and Email through GoSumo’s wizard.
Define Rules – In the admin panel, create routing rules based on intent keywords or customer properties.
Assign Agents – Add team members and set their skill tags. The system will automatically match conversations to the most suitable agent.
Monitor – Use the dashboard to track conversations, see which channel is busiest, and adjust rules on the fly.
Automate – For common queries, write canned responses that the AI can deploy instantly, freeing agents to focus on complex issues.
The Result
GoSumo’s Smart Routing turns fragmented, manual conversation handling into a cohesive, AI‑enhanced workflow. Small businesses in India can now deliver faster, more consistent customer service across all channels, improving satisfaction without adding new tools or training. The result is a leaner client‑management process that scales with growth.
Technical Highlights
NestJS + TypeScript for a strong, typed backend.
PostgreSQL + Prisma for relational data with easy migrations.
Redis + BullMQ for fast queuing and task scheduling.
Qdrant for high‑performance vector similarity searches.
OpenRouter AI for intent classification and natural language understanding.
Next.js for a responsive, server‑side rendered UI.
These components work together to provide a reliable, low‑latency routing engine that adapts to pesto business needs.
Bottom Line
GoSumo’s Smart Routing eliminates the need for multiple channel apps, reduces response times, and gives Indian SMBs a single view of all customer conversations. By combining AI intent detection with rule‑based assignment, the platform delivers consistent, automated customer service while still allowing human agents to intervene when necessary.
Live Demo: https://gosumo.aiknol.com
Source Code: https://github.com/r2st/GoSumo