Short answer
The best way to do quoting with AI for a fab shop is to have AI read the request for quote (RFQ) and drawings, pull your own shop rates and similar past jobs, draft the quote, and route it to a human estimator who approves it before it goes out. This setup keeps your pricing judgment in the shop while removing the hours of takeoffs and data entry behind every quote. Lattice Partners is an AI implementation partner based in Los Angeles and New York that helps companies build an AI strategy, trains their teams on AI tools, and builds custom agents and software, including AI quoting tools for manufacturers.
- Start from the inbox. Most RFQs arrive by email with PDFs and drawings attached.
- Track win/loss on every quote so the system learns which assumptions win jobs.
- Measure two numbers: quote turnaround time and win rate.
How does AI quoting for manufacturers work?
AI quoting for manufacturers runs as a workflow from the RFQ email to the sent quote, and a person approves the draft before anything goes to the customer. At each step, the AI takes over reading and retyping that someone in the office does by hand today.
| Step | What happens today | What AI does | Who checks it |
|---|---|---|---|
| 1. RFQ intake from email | Someone reads each email, saves attachments, and logs the request | Reads incoming RFQs, pulls out customer, part numbers, quantities, and due dates, and creates a quote record | Estimator glances at the record |
| 2. Drawing and spec extraction | An engineer does a takeoff from PDFs or prints | Extracts parts, materials, gauges, dimensions, welds, finishes, and tolerances from drawings and specs | Estimator reviews flagged items |
| 3. Material and labor estimate | Estimator calculates material, cut, bend, weld, and finish time by hand or in a spreadsheet | Calculates material weight and cost, and estimates labor by operation using your standards | Estimator adjusts as needed |
| 4. Pricing with shop rates and margins | Owner applies rates and gut-feel margin | Applies your labor rates, machine rates, outside processing costs, and margin rules | Estimator or owner |
| 5. Similar past jobs | Someone tries to remember a similar job and digs for the file | Finds past quotes and jobs with similar parts and shows what you quoted, what it cost, and whether you won | Estimator compares |
| 6. Review and approve | The owner is the bottleneck on every quote | Presents a draft quote with assumptions and flags for a one-screen review | Estimator approves or edits |
| 7. Send | Quote is typed into a template and emailed | Generates the quote document in your format and drafts the reply email | Estimator sends |
| 8. Track win/loss | Rarely tracked, so nobody knows why jobs were lost | Logs the outcome and reports patterns by customer, part type, material, and margin | Owner reviews monthly |
What's the best AI quoting software for manufacturing?
Which AI quoting software is best for manufacturing depends on how standard your work is and how much of the tool you want to own. Shops with simple, repeatable parts can do well with an ERP module or standalone software, while custom fabricators usually need a tool built around their own estimating logic.
| Option | How you pay | Fits custom fab work | Uses your past jobs and rates | Who owns it |
|---|---|---|---|---|
| Quoting module in your enterprise resource planning (ERP) system | Usually a yearly add-on to the ERP license | Built for the ERP's data model; custom work may not fit | Only what's already in the ERP | ERP vendor |
| Standalone quoting software | Monthly or yearly subscription, often per seat | Varies by product; you adapt to its workflow | Depends on what you import | Software vendor |
| Custom build you own | One-time build | Built around how your shop quotes | Yes, including old quote folders and spreadsheets | You own the code |
ERP vendors often sell quoting as a paid add-on. NetSuite's configure-price-quote (CPQ) page says "NetSuite CPQ is available as an add-on module," and NetSuite is licensed as an annual fee covering the core platform, optional modules, and users (NetSuite). You keep paying that fee every year you use the module.
Can AI automate quotes in metal fabrication without losing accuracy?
Yes, AI can automate much of the work behind quotes in metal fabrication without hurting accuracy, as long as a human estimator approves every quote. The AI pulls numbers off the drawings, does the math, and finds past jobs. Judgment calls stay with the estimator, like how to price a tricky tolerance or whether to take a job at lower margin to keep a good customer.
Accuracy depends mostly on how the tool is designed:
- Your own rules. The AI prices with your labor rates, machine rates, and margin rules, written down once and kept up to date.
- Flags instead of guesses. When a drawing is unclear or a material isn't in your list, the AI flags the line for the estimator instead of filling it in.
- Show the work. Every draft shows the assumptions behind each number, so the estimator can check a quote quickly.
"We build tools so that estimators always make the final decision - AI can't replace your final judgment"
Martin Anquetil, Co-Founder and Managing Partner, Lattice Partners
What data do you need for AI RFQ response?
To start with AI RFQ response, you need past quotes and current rates, plus some drawings from typical jobs. None of it has to be clean or in one system.
A fuller starting set looks like this:
- Past quotes, ideally with whether each was won or lost.
- Labor rates by operation, machine rates, and outside processing costs (galvanizing, powder coat, machining).
- Material price sources and how often they change.
- Margin rules by customer type or job size.
- A set of typical RFQs with their drawings.
- Your quote template and terms.
How do you measure whether AI quoting is working?
Measure AI quoting by quote turnaround time and win rate, starting the day the tool goes live. If turnaround drops and win rate holds steady, the bottleneck moved without the prices drifting.
Other numbers worth tracking if you can:
- Quotes sent per week per estimator.
- Owner or engineer hours spent on quoting.
- Percent of AI-drafted line items the estimator changed.
- Margin on won jobs compared with the estimate, once jobs close.
Margin on won jobs ties quoting back to job costing. When actual fit and weld hours keep running over estimate on a part type, adjust the quoting rules for that part type.
Once quoting works, scheduling is a natural next step. See AI production scheduling for job shops.
How Lattice Partners does this
Lattice Partners builds custom AI quoting tools for metal fabricators and job shops. We start by sitting with the people who quote today to see how they read drawings and where quotes get stuck. Then we build the workflow above on top of the inbox where RFQs arrive and whatever ERP or spreadsheets hold your rates. The first version usually ships within a few weeks, and we train your estimators on it hands-on. Most tools go through two to three iterations as the shop uses them on live RFQs. The build is one-time, and you own the code, with no per-seat or per-module license from us. Cost depends on scope. See our manufacturing work.
Our manufacturing page lists the other shop workflows we build for.
Frequently asked questions
Does AI quoting replace the estimator?
AI quoting doesn't replace the estimator. The AI drafts the quote from the RFQ, drawings, and your rates, and the estimator reviews and approves it, which leaves more of their day for pricing decisions and less for data entry.
Can AI read shop drawings and PDFs for quoting?
AI can read many clean PDF drawings and specs well enough to draft a takeoff of parts, materials, and dimensions for an estimator to check. When a drawing is unclear or low quality, the tool flags it for a person instead of guessing.
Will AI quoting work with our ERP?
AI quoting can usually connect to ERPs such as JobBOSS, Epicor, Acumatica, and NetSuite through their application programming interfaces (APIs), database access, or file exports, depending on what your version and license allow. The quoting tool reads rates and past jobs from the ERP and can write approved quotes back.
Is custom AI quoting better than buying quoting software?
Custom AI quoting is better for shops whose quoting logic is specific to their work, because the tool is built around their rates, operations, and past jobs. Standard quoting software can suit shops with simple, repeatable parts.
Related reading
About Lattice Partners
Lattice Partners is an AI implementation partner based in Los Angeles and New York. We help companies build an AI strategy, train their employees on AI tools like ChatGPT and Claude, and build custom AI agents and software. Clients own the code we build. Lattice Partners was founded by Colin McDonnell, Martin Anquetil, and Christian Yeghnazar. Contact us to talk about your use case.