Building a multi-channel AI chatbot platform for Mawjood AI

The Challenge

Mawjood AI wanted to offer their own SMB customers an AI-powered customer engagement product — but building and maintaining multi-channel chat infrastructure, an LLM-backed conversation engine, and a unified support workflow is a significant engineering commitment on its own, separate from running the rest of their business. They needed a product they could put their own brand on and sell to many customers, not a one-off internal tool.

The Approach

DevsTek built the chatbot as a multi-tenant SaaS platform designed from the start to be resold, not just used internally. Each of Mawjood AI's customers operates their own isolated instance — its own conversations, its own configuration — while running on the same underlying platform, so Mawjood AI can onboard a new customer without new engineering work per client.

The chatbot itself is LLM-powered, handling customer conversations naturally rather than following rigid decision-tree scripts, and it works the same way regardless of where the conversation starts.

The Results

• A single resellable product Mawjood AI can put in front of many SMB customers, not a bespoke build per client
• One inbox per customer regardless of which channel a conversation started on
• Routine questions handled by AI, with a clean handoff to a human when a conversation needs one
• New customers onboarded onto existing infrastructure, without new engineering per client