Questions about working with AIpplied Labs

What does a project cost?

Every project is a fixed price. The Lead engine is $3,000. The other systems run from $2,500 to $8,500 each, and the free working session turns that range into one number before anything is agreed.

Half is paid on signature and half when the system passes its acceptance test, so there is no hourly meter and no surprise at the end.

Do you charge hourly?

No. I work against a fixed scope for a fixed number, both agreed before anything starts. There is no open-ended meter running and no invoice at the end that nobody saw coming.

What does ongoing support cost?

I quote it separately from the build, and it is always optional. Most clients take it because tools are worth keeping current as the business changes, but nothing you paid for stops working if you decline it.

Do we have to commit to everything at once?

No, and it usually works better in phases. A first phase handles the single worst bottleneck, ships, and proves the approach against your real data before anything bigger gets scoped.

Each phase is priced and agreed on its own. Stopping after one is a normal outcome.

Is the working session really free?

Yes. Twenty minutes, no slides, and no obligation to go further. We work on your biggest time sink live, and the roadmap or demo you leave with is yours whether or not anything gets built.

Are we locked in to anything?

No. You own the code and the documentation outright, nothing is licensed back to you, and nothing stops working if you decide not to continue. I offer support because it is often useful, never because the work depends on it.

Who owns what gets built?

You do, outright, including the code and the documentation. Nothing is licensed back to you and nothing stops working if you decide not to continue. Ongoing support is available because it is often useful, never because the work depends on it.

If you reuse parts of our build for other clients, what is ours?

Your data and your settings are yours permanently. Nobody else sees them, and nothing specific to how your business runs gets carried into somebody else's build.

I keep the generic building blocks and reuse them across projects. That reuse is a large part of why a build takes weeks instead of months, and you don't pay to reinvent plumbing that already works.

What if it doesn't work?

Nothing ships until you have tested it and are comfortable with how it works, so a system that can't do the job never reaches handover. I work in weekly increments for the same reason: problems surface in week one, while they are still cheap.

What if we aren't sure yet what we need?

That is exactly what the free working session is for. Bring the biggest time sinks and the parts of the week that feel heaviest, and the outcome is a straight answer about whether AI helps, what it would actually look like, and what isn't worth building. No pressure and no commitment attached.

What does the process look like?

Three steps. Working session: 20 minutes on your biggest time sink, and you leave with a one-page roadmap or a quick demo. Build: weekly increments against a fixed scope, and I write the code myself after running the session. Handover: owned code with documentation written for the team that inherits it.

Support continues by arrangement, never as a condition of keeping the system running.

How long does it take to build?

Most projects are ready two to four weeks after the discovery call. More complex integrations take longer. Nothing ships until you have tested it and are comfortable with how it works.

When could you start?

The working session is usually the same or the following week. Build slots depend on what I already have running, and the session ends with a real date on it.

How much of our time will this take?

About twenty minutes for the working session, then your honest feedback once there is something to look at. Progress shows up in weekly increments. Nothing to chase, nobody to manage, and no status meeting to sit through.

Does anyone on our team need to be technical?

No. You describe the problem in plain language; I handle everything technical. The only real asks are about twenty minutes during discovery and your honest feedback once there is something to look at.

Will our team need training?

Very little, by design. Tools that need a training program to survive tend to get abandoned. Handover includes documentation written for whoever inherits the system, in the language they already use for the work.

Will it work with the tools we already pay for?

That is the default. Part of the working session is finding out which tools are already paid for and underused, because the cheapest system is usually the one that runs on what you already pay for.

Our records are a mess. Is that a problem?

It is the normal starting condition. Most of the useful material is already there, scattered across a shared drive, an inbox, a chat history, and a few people's heads.

I spend part of the first week of the build finding out what shape it is actually in, which is why I don't quote a fixed system before that look.

Where does our data live and who can see it?

In your accounts, under your control, on infrastructure you own. I need access to build, and I hand every credential back at handover.

I scope anything sensitive in the working session before writing a line of code, because getting that wrong is expensive.

What happens after launch?

You are not handed a finished product and left with it. Ongoing management is offered so tools stay updated and keep improving as the business changes, and most clients take it. It stays optional, and the work is yours either way.

Who keeps a company brain up to date after handover?

Somebody on your side has to own it, and the agreement names who before the build starts. A knowledge system nobody curates goes stale in months, and a stale answer reads to staff like the whole thing is broken.

The handover includes how to add to it and what to review, and I can hold that job under ongoing support if you would rather I kept it.

How do we get started?

The form at the bottom of this page. I reply within one business day with two times to choose from. Then we have a 20-minute working session on your biggest time sink, and you get a one-page roadmap or a quick demo the same day.

No newsletter and no follow-up sequence. One reply, from me.

What kinds of things do you actually build?

Six areas. Websites built from scratch for one business, never dragged out of a template. Getting found in both search and AI assistants. A company brain that gathers scattered knowledge into one place staff can question in plain language. Chatbots and agents that answer and triage. Systems for winning and keeping clients. And taking the administrative weight off the week.

Every engagement starts as a conversation and ends as working software you own.

Can you build something specific to the way we work?

Yes, and most of my work is exactly that. The systems described on this page are patterns rather than products. I build each one against how one business actually runs, using the tools it already pays for.

If the thing you need isn't listed anywhere here, describe it in the working session. I'll give you a straight yes or no, with the reason attached.

Do you build websites?

Yes. I build each one from scratch for one business, with no template underneath. Fast on every device, shipped in weeks, and handed over as code you keep for good.

Can you get us found in ChatGPT and AI search?

Yes. People don't only search Google anymore; they ask an assistant who to hire. I build and structure sites so both kinds of search can find a business, understand it, and recommend it by name.

One client's first result from it was a customer two hours away who said she found them using ChatGPT.

Can you stop inquiries from going cold?

Yes. Every inquiry lands in one place the moment it arrives, tagged by where it came from, with a first reply already sent. Follow-up reminders fire on their own, so nothing goes quiet without someone hearing about it.

Can you build a chatbot that answers like someone who works here?

Yes, and that is the point of it. The AI teammate answers on the site and in the inbox from your own material, handles what it can, books what it is allowed to, and hands over what needs a person.

It escalates the moment it isn't sure, which is the part that makes it safe to put in front of clients.

What is a Company Brain?

One place holding the documents, the decisions, and the current state of the work, answerable in plain language. Ask it about the warranty on installs and it answers from your own warranty terms and tells you the page it came from.

It is also the layer every other AI tool builds on, because it is what lets them answer from your actual business.

Can you fix how we onboard new clients?

Yes. A new client's first month can run on rails: the welcome, the intake, the access, and the kickoff each fire off the step before them, so nobody has to remember them. Six steps, none of them on anyone's to-do list.

Can clients see where their job stands without asking?

Yes. The delivery line puts the same path from signed to delivered on one page, each stage dated and visible to both sides. The client sees that page too, which is what removes the status meeting.

Can you automate our monthly reporting?

Yes. It gets built one time and then builds itself. The numbers pull straight from the systems that produce them and it arrives on the same day every month, with nobody touching it.

Why not just use ChatGPT ourselves?

A chat window answers when somebody remembers to open it, starts from nothing every time, and never fires on its own at nine at night when an inquiry lands.

My work is turning the useful part into something that runs without anyone driving it, with your own documents behind it and a clear line for when it should hand over to a person.

Is this going to replace people on our team?

No. The point is to take the repetitive, low-value work off their desks so they can spend their time on the parts that genuinely need judgment. The businesses that get the most out of AI use it to make a small team more effective.

Will you tell us if something isn't worth building?

Yes, and the working session is designed to produce exactly that. You leave with a one-page roadmap that names what is worth building and what is not, and the second part is often the more useful one.

I spent years watching organizations spend millions on technology that could have shipped in weeks. That is the reason the list exists.

Is there anything you refuse to build?

Anything that needs a person to babysit it to stay useful, anything that replaces a judgment call that should stay with a human, and anything where the honest answer is that a spreadsheet already does the job.

I name those in the working session, in writing, before there is a proposal to argue about.

What kind of business is this for?

Businesses small enough that the founder still feels every wasted hour, and established enough that the same work repeats every week. The pattern is always the same: a report rebuilt by hand every month, inquiries that land at nine at night and get answered at nine the next morning, an answer that exists in a document nobody can find.

None of that calls for a bigger team. It calls for the right tool, scoped properly, and built for how one business actually works.

Is my business too small for this?

Probably not. Running like a modern operation shouldn't take an enterprise budget, and that is the whole reason my company exists. I scope the work to the problem it solves, with no minimum engagement size.

If the working session finds nothing worth building, I tell you that, free, and nobody wastes the money.

Which industries do you work in?

My work isn't tied to one industry. What repeats across all of them is the shape of the problem: knowledge stuck in people's heads, inquiries going cold, and the same report rebuilt by hand. My background spans public service, applied research, and large private technology programs, plus a portfolio of my company's own ventures.

Do you work with businesses outside Chicago?

Yes. Everything here can be built remotely. I'm based in the Chicago area and work with clients anywhere. Discovery, check-ins, and demos all happen over video.

Who is behind AIpplied Labs?

I'm Kevin Magnan, and AIpplied Labs is my company. I started out as a police officer, moved into applied research, and then spent 10 years in management consulting, with more than $50 million in engagements before I started the company.

None of it began in computer science. It began in work where getting something wrong had consequences for real people, and the recurring lesson across all of it was that budget never guaranteed the result.

Who actually does the work, and how big is the team?

I do. I scope it, I run the working session, and I write the code that ships. That removes the handoff where most agency work quietly loses its detail.

My company is effectively one person, and that is the offer: no account managers, no handoff chain. I bring in a specialist when a build genuinely needs one, and nobody gets added to a project to fill out an org chart.

How is this different from hiring an agency?

No account managers and no handoff chain. You talk directly to me, and I write the code, which makes the work faster and removes the layer where detail usually gets lost. It also costs less, because there is no agency overhead priced into it.

How many clients do you take at once?

Few enough that yours isn't waiting behind three others. Build slots are the real constraint, which is why the working session ends with a date.

If I don't have a slot when you need one, I tell you that in the working session.

Do you build your own products, or only client work?

Both. I run my own portfolio: PackLlama for moving inventory, Doughby for sourdough baking, and AIppliance Manager on iOS.

They exist because a company that only builds for other people stops learning what actually ships. Everything I use on a client project, I've run in production myself first.

Do you use this kind of system yourself?

Every day, and that is where most of my opinions come from. I run my own knowledge base, my own automated reporting, and my own agents against real work.

It also means the failure modes I bring up in the working session are ones I've already hit myself.