Turn mill test reports into a material heat number traceability log with AI

Short answer

Use Claude or ChatGPT, then Copilot in Excel: The AI reads mill test report PDFs into a fixed register (heat, product, size, grade as printed, PO, certificate, file and page, UNREADABLE instead of guesses), then Copilot matches heats to receiving and the cut list and flags missing certificates, orphan certificates, near-miss heat numbers and grade or size differences from the PO. Shops that have to keep heat and mill test report (MTR) identity through fabrication often do it with a binder of certificates and a spreadsheet typed by hand, and the gaps only show up when someone asks for the paperwork. Have Claude or ChatGPT read each MTR PDF into a fixed table (heat number, product, size, grade as printed, PO and certificate number), then use Copilot in Excel to match those heats to your receiving records and cut list and flag pieces with no certificate on file. Checking certificates against the purchase order and the material specification stays with your QC person. A person still checks: every exception against the steel's physical marking and the certificate, and the review of each MTR against the PO and material specification by QC staff.

AI tie-in

Tool: Claude or ChatGPT, then Copilot in Excel

The AI reads mill test report PDFs into a fixed register (heat, product, size, grade as printed, PO, certificate, file and page, UNREADABLE instead of guesses), then Copilot matches heats to receiving and the cut list and flags missing certificates, orphan certificates, near-miss heat numbers and grade or size differences from the PO.

Prompt 1

Read these mill test reports [upload PDFs]. Build a table with one row per heat number per certificate: file name, page, certificate number, mill or supplier as printed, heat number, product and size, grade or specification as printed, quantity or weight if shown, and the customer PO number if shown. Copy every value exactly as printed. Write UNREADABLE for anything you cannot read and NOT SHOWN for anything missing. Do not evaluate whether any chemistry or mechanical value meets a specification.

Prompt 2

Here is my MTR register [paste], my receiving log: PO, PO line item, size, grade ordered, heat number as marked on the steel, date received [paste], and my cut list: piece mark, heat used, quantity [paste]. Match on heat number and give me four lists: heats received with no MTR, MTRs with no matching receipt, heat numbers that differ by one or two characters from another heat (possible typos), and rows where size or grade on the MTR differs from the PO line. Then list piece marks with no heat recorded. Use only these tables and do not judge material conformance.

Human check: Every exception against the steel's physical marking and the certificate, and the review of each MTR against the PO and material specification by QC staff.

The problem

Turn mill test reports into a material heat number traceability log with AI. This is a common, time-consuming task for a fabricator.

What this workflow does

Shops that have to keep heat and mill test report (MTR) identity through fabrication often do it with a binder of certificates and a spreadsheet typed by hand, and the gaps only show up when someone asks for the paperwork. Have Claude or ChatGPT read each MTR PDF into a fixed table (heat number, product, size, grade as printed, PO and certificate number), then use Copilot in Excel to match those heats to your receiving records and cut list and flag pieces with no certificate on file. Checking certificates against the purchase order and the material specification stays with your QC person.

How it works: AI drafts or analyzes. A person operates the tools, checks the work, and takes action.

Your checkpoint: Extraction and matching only. Models misread heat numbers and multi-heat certificates, so every row is checked against the PDF and the steel. Whether material conforms to the order and specification, and what traceability the contract requires, are decisions for your QC staff and the contract documents.

Before you start

Access to the named tool; confirm license, version and cost with the vendor.
A small, non-confidential sample to try the workflow on first.
A qualified person to review the output before it is used.

Tools and access

Follow the steps

Decide what level of traceability the contract actually asks for (lot, piece mark or piece, main members only or all material) before building anything. The AISC Code of Standard Practice, Section 6.1, separates material identification from traceability; read it and your contract documents in your own licensed copies.
Set up one workbook per job with four tabs: MTR register (one row per heat per certificate), receiving (PO, delivery ticket, item, heat number as marked on the steel, date, who checked it), cut and issue (piece mark, heat used, quantity, remnant and its re-marked heat), and an exceptions tab.
Upload a batch of MTR PDFs to Claude or ChatGPT and have it fill the MTR register columns exactly as printed, with the file name and page for every row. Anything it cannot read is UNREADABLE, never a guess, and it does not judge whether the chemistry or mechanical results meet any specification.
Paste the register and the receiving tab into Copilot in Excel and have it match on heat number. Flag heats received with no MTR on file, MTRs with no matching receipt, heat numbers that differ by a character or two (likely typos), and grade or size on the MTR that differs from the PO line.
Join the cut list to the receiving tab so every piece mark shows its heat and certificate, and list piece marks with no heat recorded. Remnants carry their original heat forward.
Your QC person checks every exception against the physical marking on the steel and the certificate, reviews the MTRs against the purchase order and the material specification, and signs off. Re-run the matching whenever new certificates or receipts come in.

Try a small example

Replace the bracketed fields with your own project details. No tested sample run is published for this workflow yet.

Sample input

Fill in the bracketed [fields] in the text below with your project data.

Starter prompt

Read these mill test reports [upload PDFs]. Build a table with one row per heat number per certificate: file name, page, certificate number, mill or supplier as printed, heat number, product and size, grade or specification as printed, quantity or weight if shown, and the customer PO number if shown. Copy every value exactly as printed. Write UNREADABLE for anything you cannot read and NOT SHOWN for anything missing. Do not evaluate whether any chemistry or mechanical value meets a specification.

Expected output

Not published yet. Treat any output as a draft and check it against the source documents.

Check the result

Every value, citation or quantity traces back to a source document.
Missing information is marked as missing, not filled in.

Limits and recovery

Extraction and matching only. Models misread heat numbers and multi-heat certificates, so every row is checked against the PDF and the steel. Whether material conforms to the order and specification, and what traceability the contract requires, are decisions for your QC staff and the contract documents.
Check your organization’s data policy before sharing project files with third-party tools.

Stop or switch when: The source cannot be verified, permissions are missing, or the result contains unsupported facts.

Frequently asked questions

How does AI help with this?

Claude or ChatGPT, then Copilot in Excel: The AI reads mill test report PDFs into a fixed register (heat, product, size, grade as printed, PO, certificate, file and page, UNREADABLE instead of guesses), then Copilot matches heats to receiving and the cut list and flags missing certificates, orphan certificates, near-miss heat numbers and grade or size differences from the PO. A person still checks: Every exception against the steel's physical marking and the certificate, and the review of each MTR against the PO and material specification by QC staff.

How do I turn mill test reports into a material heat number traceability log with AI?

Decide what level of traceability the contract actually asks for (lot, piece mark or piece, main members only or all material) before building anything. The AISC Code of Standard Practice, Section 6.1, separates material identification from traceability; read it and your contract documents in your own licensed copies. Set up one workbook per job with four tabs: MTR register (one row per heat per certificate), receiving (PO, delivery ticket, item, heat number as marked on the steel, date, who checked it), cut and issue (piece mark, heat used, quantity, remnant and its re-marked heat), and an exceptions tab. Upload a batch of MTR PDFs to Claude or ChatGPT and have it fill the MTR register columns exactly as printed, with the file name and page for every row. Anything it cannot read is UNREADABLE, never a guess, and it does not judge whether the chemistry or mechanical results meet any specification. The full step list is above.

Does this use an AI agent or regular software?

It is labeled AI assists: AI drafts or analyzes. A person operates the tools, checks the work, and takes action.

What tools do I need?

Claude, ChatGPT, Microsoft Excel, Microsoft Copilot. Confirm license, version and cost with the vendor, and try it on a small non-confidential sample first.

Has this workflow been tested?

Workflow guide · not independently tested. Written from vendor documentation and practitioner use. Labels apply to the steps described here, not every feature of a product. Links were checked when the page was written.

Evidence and maintenance

Written from vendor documentation and practitioner use. Labels apply to the steps described here, not every feature of a product. Links were checked when the page was written.