XecSuite
AI operating layer for supply chain teams
A governed workforce of AI agents that runs the back office of a logistics company, with a human approving what matters.
- 6
- connected modules
- 8
- agent roles
- 2
- countries, Canada and the US
Problem
Freight, finance and customer work in a logistics company lives in email, spreadsheets, a TMS and people's heads. Every request means someone reading, deciding and retyping, and the knowledge of how things are done leaves with them.
What I built
- A tenant-private company memory that connects the client's systems and knowledge once.
- A support agent that reads, answers and resolves tickets on its own.
- A voice agent, a system that places phone calls, and a meeting agent that joins, tracks and follows up on meetings.
- Email automation and custom per-client features on top of the core.
- Approval gates: every agent action that matters waits for a person.
How it works
Modules around the memory core: Freight Desk, Load Coverage, Logistics, Premium CRM, Meeting Agent and RFP. Hybrid model routing sends routine steps to private models and high-stakes reasoning to frontier models. Built with Nik, my co-founder.
Screens
Results
Built and awarded at a Startup Weekend; now live, and improved every week with new custom features.
My role
Co-founder and CTO: architecture, agent runtime, model routing and product engineering.
Stack
- / Next.js
- / TypeScript
- / Agents
- / Hybrid model routing
- / Voice