Korso Turns Manufacturing RFQs Into Quotes Automatically Because Nobody Should Spend Three Days on a Spreadsheet

The Macro: Manufacturers Are Leaving Money on the Table Every Day

Here is how quoting works at most small and mid-sized manufacturers. An RFQ (request for quote) arrives by email. Someone opens it, reads through the line items, pulls up historical pricing, checks material costs, calculates labor time, and manually builds a quote in a spreadsheet or Word document. This process takes hours to days depending on complexity. During that time, the customer might be getting faster quotes from competitors.

The speed of quoting directly affects revenue. Studies show that the first supplier to respond to an RFQ wins the business 30-50% of the time. Manufacturers who take three days to quote are losing to competitors who take three hours. But building quotes faster with manual processes means cutting corners on accuracy, which leads to underpricing, overcommitting, or embarrassing errors in front of customers.

The RFQ process is just the beginning. After the quote is accepted, purchase orders need to be tracked through production. Supplier communications need to be managed. Documents need to be classified and filed. All of this is done manually in most shops, often by the owner or a small office staff who are already overwhelmed.

Korso, backed by Y Combinator, is building what they call the intelligence layer for manufacturing, automating the full quote-to-order workflow with AI.

The Micro: From Email to Quote to PO, Automated

Korso handles the end-to-end process. RFQs arrive via email or WhatsApp. AI extracts line items, quantities, specifications, and delivery requirements from the documents. The system cross-references historical pricing, material costs, and customer relationships to generate a professional quote. Purchase orders are tracked through production with automated status updates.

The multi-channel ingestion is practical. Manufacturers get requests through email, WhatsApp, and sometimes phone calls. Meeting customers where they communicate rather than forcing them into a portal reduces friction.

The RAG-powered AI assistant understands the manufacturer’s specific business: their pricing history, their capabilities, their customer relationships. This context makes quotes more accurate because the system is not just running a generic calculation. It is learning from every previous quote and order.

The founding team includes Daichi Hiraoka, Alex Liu, and Martin Pan. Martin brings prior experience at GM, which provides direct exposure to manufacturing operations.

The competitive space includes ERP systems like JobBoss and Epicor that have quoting modules, but these are clunky and require extensive manual input. Dedicated quoting tools like Quoter and PandaDoc are too generic for manufacturing. Arzana, also in this YC batch, targets a similar market with their Office Execution System. The overlap between Korso and Arzana is significant, and it will be interesting to see how they differentiate.

The Verdict

Manufacturing operations are dramatically underserved by modern software. Most shops run on a combination of ERP systems from the 2000s, spreadsheets, and paper. The opportunity for AI to transform these workflows is substantial.

At 30 days: how many quotes per day are manufacturers generating through Korso, and how does the turnaround time compare to their manual process? The speed improvement is the headline metric.

At 60 days: are win rates improving for manufacturers using Korso? If faster quoting leads to more won business, the ROI is immediate and obvious.

At 90 days: how many manufacturers are using Korso for the full quote-to-order workflow versus just quoting? Expanding from quoting into PO tracking and production management deepens the product’s value and increases switching costs.

I like this market. Manufacturers are practical buyers who adopt tools that make them money. If Korso demonstrably helps them quote faster and win more business, word will spread quickly through manufacturing networks. The key is getting the first cohort of manufacturers to see results.