Software and AI companies are discovering what people in the software world call "boring businesses": manufacturing, distribution, construction. Your inbox is full, the paperwork never ends, and the pitch sounds great. Expect a lot more sales meetings.

Most of those vendors will not tell you how the work gets done, and you should know what you are paying for.

1. Software got cheap, and most vendors won't say so

For decades, people wrote every line of custom software by hand. That is why implementation projects carried six- and seven-figure price tags and ran for a year.

That has changed. The companies that build AI now use it to write most of their own code.

75%of all new code at Google is AI-generated and approved by engineers, up from 50% the fall beforeSundar Pichai, Google Cloud Next, April 2026
80%+of the code merged into Anthropic's codebase was authored by Claude, as of May 2026Anthropic
0lines of hand-written code in a million-line internal product an OpenAI team built with its Codex agentOpenAI, 2026

A system built around how your company works is no longer a luxury that only a Fortune 500 company can afford. If a vendor prices and schedules as if it were 2019, ask why.

2. A demo takes about a day

Say your company is a good fit. A team can build a working demo on your own documents in about a day.

A good partner will not hand it to you after a day, and will not pretend it took a month either. In our view, some vendors stretch the timeline because a longer wait makes the price feel earned and softens buyer's remorse. The demo is the easy part, and writing the software is the fast part. The rest of the project takes longer.

3. The real cost is getting clear on your work

You will meet two kinds of vendors. The first says software is cheap now, so they can do everything quickly and cheaply. Be careful with them. Cheap code is real, but it only covers a fraction of the job.

The second kind says the work costs real money and tells you why. Your money pays for time spent getting clear on your work.

Clarity means naming each request that comes in and the decisions people make on each one. It also means marking the spot where a person has to check, and writing down what happens when the answer is unclear.

That clarity lives in people's heads and on paper. Getting it out takes weeks of sitting with the people who do the job. The partner you want works on proof over promise, which means they show you results on your own documents.

Today a builder can build almost anything and copy almost any software, so a good partner helps you decide what to leave out. They look for steps you can remove before you automate anything, and they cut features you do not need. They also ask which reports exist only because someone who left ten years ago asked for them.

In our view, you should settle three questions before you build.

  1. What are we really optimizing for? Faster quotes, fewer certificate errors, less time lost in the inbox. Pick one first.
  2. What are the constraints? Which systems the AI may touch, what it may never do, and what needs a signature. More constraints make a safer system.
  3. What are the safety mechanisms? How the AI checks its own work, what it flags, and who reviews what.

4. AI will make mistakes, so design where they land

AI will not get it right 100% of the time, no matter who sells it. Anyone who promises 100% accuracy is selling.

Ask how and where it fails. If the AI makes a mistake inside its own workspace, in a draft that a person reviews before anything goes out, the mistake costs a minute. If it makes the same mistake while acting directly in production, sending an email, posting to the ERP, or releasing a shipment, you find out from your customer.

A fully automated AI system is easy to market, and we do not recommend one. Look for a system that does the first pass on the work. It processes your files and connects information you would otherwise hunt for. It turns the paperwork into digital records you can see, and it tells you when a job needs your attention.

A good system also checks each draft against your records, flags what it is unsure about, and hands the draft to a person who can check it and sign. The work sits in your inbox and your files, where your team already works.

A system like this can also change as fast as you do. When your process changes, you describe the change, and AI builds it into the system. A person tests each change before it goes live. Your team skips the vendor call and the weeks of waiting for a small edit.

5. Where custom software still costs real money

Some costs stay high even with cheap code, and they are the ones that protect you.

Security and compliance are the big ones. If a vendor handles your data, ask whether they have a SOC 2 report.

A SOC 2 report is an independent examination by a CPA firm. The firm checks how the company controls the security, availability, and confidentiality of the information it handles, and it watches those controls for months before it issues the report.

Some data needs tighter rules than other data. Drawings, customer prices, and export-controlled work each need their own rules about who and what can see them.

We think those costs are worth paying. They are the difference between a tool and a system you can trust with your customers' information.

6. The questions to ask any AI vendor

Bring these to your next vendor meeting. The answers will tell you quickly who has done this before.

AskA good answerA red flag
Can you show it working on our documents, not a demo set?Yes, send us a few real examples.Only their own polished demo.
How much of your cost is software, and how much is people and process?A plain breakdown, with most of the effort in understanding your work."The platform does it all."
What will you recommend we not build?Specific things to cut or simplify.Nothing. Everything is in scope.
How will you test it before it touches live work?Your past cases with known answers, graded against a bar you agree on in writing."It's very accurate."
What happens when the AI is unsure?It flags it and stops, and a person decides.It picks the most likely answer.
What can it do without a person approving?Only steps that stay inside the building.Sending, posting, or releasing on its own.
Where does our data go, and is it used to train anyone's model?Named providers, and a written no on training.Vague answers or "industry standard."
Do you have a SOC 2 report, or another independent security review?Yes, and here it is, or a clear timeline and the controls in place today."We take security seriously."
Can we see a log of everything the AI did?Yes, step by step, with times and sources.No log, or only for their engineers.
How do we change it when our process changes?Describe the change, it is built, tested, and approved by a person.A change request and a new quote every time.
What happens if we leave?You keep your data, your test cases, and your records.Your data is locked in their format.
Who from our side needs to be involved, and for how long?Named roles and hours, including leadership."You won't need to do anything."

What to do this week

  1. Print the table above and bring it to your next AI vendor meeting. Ask every question, and write down the answers word for word.
  2. Pick three real documents of your own, such as an RFQ, a certificate request, and a messy email, and ask each vendor to run them through their system.
  3. List what you would want the AI to never do on its own. If a vendor's product cannot respect that list, it is the wrong product.
  4. Read Software Got Cheap, Clarity Didn't: How to Define a Job Before Any AI Touches It for what clarity looks like in practice. Use the free AI Employee Handbook to learn how to test an AI before it goes live.

Questions people ask

How much does custom AI software cost for a manufacturer?

The code itself costs far less than it used to. Most of the project cost pays for people to learn your process, decide what to leave out, and test the system on your own documents. Security and compliance work adds cost too.

Should we buy a fully automated AI system?

We do not recommend it. AI will make mistakes with any vendor. The safer design keeps the AI in its own workspace for the first pass. A person approves anything that reaches a customer, the ERP, or a shipment.

Terms in this piece

SOC 2
an independent examination by a CPA firm of how a service company controls the security, availability, processing integrity, confidentiality, and privacy of the information it handles.
Demo
a working example built to show what a system could do. Useful only when it runs on your own documents.
Fully automated
a system that acts on its own without a person approving the output. We recommend against it for anything that reaches a customer.
AI-enabled system
a system where AI does the first pass, flags what it is unsure about, and a person verifies and signs.
Proof over promise
judging a vendor by results on your own documents instead of by claims.