The situation
Rivus was operating with disconnected systems: Odoo, VTEX, email automation through Microsoft 365, and lead scoring were all siloed. Sales reps copied data between platforms by hand, and the work that actually moves a deal, building a quote and turning it into an invoice, was manual and slow, taking hours per order. Inbound arrived faster than anyone could answer it, so inquiries went cold waiting on a reply, and leadership had no real-time view of pipeline or revenue.
What we built
We embedded with the Rivus team and built it hands-on. Underneath sits a unified data layer across Odoo, VTEX and Microsoft 365, a retrieval-grounded knowledge base over the product catalog so an answer can always be traced back to a source, and a knowledge graph that connects a customer and a part number through to the quote itself. On top of that runs Carlos, an AI sales operating system that now takes first contact on every inbound inquiry: it answers from the knowledge base, qualifies and routes the lead, and carries it through to an automated quote in Odoo without a rep retyping anything. We did the implementation ourselves, trained the team on it, and stayed embedded until it was running in production.
- Carlos takes first contact on every inbound inquiry, answered inside five minutes
- Answers grounded in a retrieval knowledge base over the product catalog, traceable to source
- A knowledge graph ties customer and part number through to the quote
- Automated quoting in Odoo, from inquiry to quote without a rep retyping
- Connected Odoo, VTEX and Microsoft 365 into one data layer
- Real-time pipeline and revenue dashboards for leadership
Built on Odoo, VTEX, Email automation via Microsoft 365 Graph, RAG knowledge base, Knowledge graph, Automated quoting.
What changed
Carlos now fields every inquiry Rivus receives, more than 2,000 a month, and answers each one inside five minutes instead of leaving it to sit. Quoting and invoicing went from hours to minutes, which lifted team productivity by 30% and freed reps to sell rather than retype. Booked calls rose 50% in the first 90 days, and the system has since brought in over $12 million in revenue. Leadership has a real-time view of pipeline for the first time.
