Concept 03
Turning a two-week quoting process into a same-day answer.
- Sector
- Industrial manufacturing
- Disciplines
- AI & Automation, Software, IT Consulting
- Year
- 2025
- Status
- Concept project
The premise
A contract manufacturer receives RFQs as PDFs, CAD files and forwarded email chains. An estimator interprets them by hand. Quoting takes ten to fourteen days, the company only bids on what it can process, and every estimator prices slightly differently.
How we would approach it
- 01
Extract before you estimate
A document pipeline reads the RFQ set — drawings, tolerances, quantities, terms — into structured parameters, flagging anything ambiguous instead of guessing.
- 02
Make pricing logic explicit
The rules estimators hold informally become a versioned, testable model. Disagreements surface as data rather than as differences between people.
- 03
Keep the human on the decision
The system produces a draft quote with its assumptions visible and its confidence stated. The estimator reviews and signs. Nothing goes out unattended.
- 04
Close the loop on win/loss
Outcomes feed back into the model so pricing improves against reality, not against opinion.
What we would build
RFQ extraction pipeline with ambiguity detection
Versioned pricing engine and test suite
Estimator review console with assumption trail
Win/loss feedback and reporting
What it is meant to prove
Concept goal: compress quote turnaround enough to change which opportunities the business can pursue, while making pricing consistent and reviewable.
Stated as an objective, not an outcome. This concept has not been delivered for a client, so there is no measured result to report — and we will not invent one.
Next concept
Signal
Start a project
A first conversation is a conversation, not a pitch. Bring the problem in whatever shape it is currently in — we will tell you honestly whether we are the right people for it.
- Typical first engagement
- 2–3 weeks
- Working model
- Embedded, in your tools
- Reporting
- Weekly demo, monthly review
- Handover
- Documented, always