How Mattsco Cut RFQ Processing Time by 70% With AI Automation
Replacing manual procurement intake with an AI-powered RFQ platform built on Azure
How Mattsco Cut RFQ Processing Time by 70% With AI Automation
Replacing manual procurement intake with an AI-powered RFQ platform built on Azure
Outcomes at a Glance
Manual RFQ processing time cut sharply
70% reduction
Any customer RFQ format handled without manual entry
AI parser reads Excel, PDF, plain text, and custom templates
Vendor coordination removed from the sales team's plate
Outreach and follow-ups run automatically
Full platform ownership with no vendor dependency
Complete codebase delivered to Mattsco
About the Client
The Situation
Mattsco is a US-based distributor of industrial hardware: pipes, flanges, fittings, fasteners, and filters, supplied to a wide customer base with high quote volumes and tight turnaround expectations. Their business runs on RFQs. A customer sends a request, Mattsco prices it against internal stock and third-party vendors, and sends back a competitive quote. Speed and accuracy are what keep customers coming back. A slow response or a pricing error means a lost order.
The problem was the RFQs themselves. Customers sent them in whatever format suited them: Excel spreadsheets with varying column structures, tabular PDFs, plain email text, templates unique to each account. No two looked the same. Each one had to be read, interpreted, and manually entered into Mattsco's system before anyone could start on pricing.
On top of that, the quoting process involved two pricing sources: internal inventory and third-party vendors. Vendor outreach was done by hand, follow-ups were chased individually, and consolidating everything into a final quote was a separate task again. There was no central view of where any RFQ stood at any point in the process.
The Impact
The manual process created delays at every stage. The gap between a customer sending an RFQ and receiving a quote was longer than it needed to be, leaving room for faster competitors. Inconsistent input formats meant interpretation errors made it into quotes, creating rework and eroding customer confidence. And because the sales team was spending considerable time on data entry and vendor follow-up, there was less capacity for the customer-facing work that actually drives revenue.
For a business where quote volume and conversion are directly tied, every hour lost in the process had a cost.
The Resolution
In 2023, when LLM-based document parsing was still early in the industry, Edstem built Mattsco an AI-powered RFQ platform that replaced the manual intake process from end to end.
The centrepiece was an LLM-driven email parser, built in Python, that reads incoming RFQs in any format: Excel, PDF, plain text, or custom templates. It identifies item names, quantities, and units, then maps them directly to Mattsco's internal product database. What had been a manual data entry task across dozens of different layouts became fully automatic.
From there, the platform handles the full workflow. Vendor outreach goes out automatically through SendGrid, with follow-up emails triggered by response timelines. As vendor quotes arrive, the platform consolidates them with internal inventory pricing and generates the final customer quote. A central dashboard gives the sales team and management a live view of every RFQ, from the moment the email lands to the point the invoice goes out.
Processing time fell by 70%. Quotes go out faster and with fewer errors. Vendor coordination runs without the sales team having to chase it.
The platform is built on Microsoft Azure with serverless Function Apps handling integrations, on an architecture designed to grow alongside the business. Mattsco owns the full codebase outright, with no ongoing dependency on Edstem or any third party for access or modifications.
Ready to automate your RFQ or procurement workflow?
If your team is spending time on manual quoting, data entry, or vendor follow-up, Edstem can help you build a solution that takes it off their plate.
MORE CASE STUDIES
