For industries that need more efficiency.

Today everyone can build with AI. I find, with your team and your customer, which problem is worth solving, build the first version, and we measure whether they use it before you spend months on that operation. The gain is efficiency: AI handles what repeats and your company grows.

Agendar conversa de 30 minutos See what I've already built

Is this your moment?

If two or more are part of the routine, the problem is not a lack of effort.It's just that the right project, in the right size, still doesn't fit.

The cost of building before knowing what it's worth building.

  • Months of team time on something the customer does not use.

  • A product that another AI starts to make for free.

  • Projeto adiado que continua custando.

  • Gente boa presa no trabalho repetido.

With AI, building is easy.The difficult thing is to build what another AI cannot replace tomorrow.

Efficiency: brute force or intelligence.

In a scenario of uncertainty, growing through brute force, hiring more people, weighs heavily.The other solution is intelligence: doing more with what the company already has.

AI is now good enough to read documents, understand orders and fill out systems.But many AI projects stop in the middle, and many things that are built today, another AI does tomorrow.What lasts depends on your data, your context and your way of deciding.So what separates the ones that work is starting small and measuring usage.

From the problem to a first version that your team and your client use.

  1. 01

    Descobrir o que vale resolver.

    I listen to your team and your customers to find the problem at the top of the list.We choose a problem, a user and a goal.If a general AI already solves it, we don't build it.

  2. 02

    First version.

    I build it with AI, so you can see it running, and not just in a document.It's small on purpose: it's for your client to actually use and show what works.

  3. 03

    Medir o uso.

    We define what value is for your customer and monitor activation, time to first value, retention and who becomes a frequent user.These are the numbers that tell you the next step.

  4. 04

    Decidir.

    The decision has criteria: move forward, adjust course, stop or collect more data before investing more time and resources.And everything we learned is documented.

Products I've already built.

Each product below came from a real problem and was built with AI.These are not industry cases yet.

ThesisOne

It started as an experiment spreadsheet. Today it records hypotheses, documents what each test taught, and is gaining a decision matrix: advance, pivot, stop, or test with more data. That is what a general AI does not have: your team's history and decision criteria. It keeps evolving with use.

Conhecer o ThesisOne

  • Marketplace de arquitetura comercial:

    two-sided platform that connects accredited architects to business owners.I made the first MVP, and then a professional development team built the complete system.

  • Frota:

    mileage and fueling entered via cell phone, with a photo of the receipt, instead of a spreadsheet.

  • Bids:

    a radar that searches the PNCP, filters by keyword and uses AI to classify IT, telecom, network and digital security hires.

I also built Buskai, a directory of AI tools for Latin America, and an AI-powered CRM for prospecting.

Escopo e valor, definidos juntos.

Before understanding its operation, any number would be a guess.I start from what the problem costs today in time and resources, and the value of the first version is measured against that.I close scope and value with you before starting.

Agendar conversa de 30 minutos

Questions

What if ChatGPT or Claude already does this?

They do a lot of things, and another AI could replace one of your products soon.That's why I only propose building where the value is in what a general AI doesn't have: the data, the context and your company's way of deciding.If a generic AI solves it, I'll say it right at the beginning.

Isso substitui o meu time?

No. The AI ​​takes over what is repeated, and your team continues with what is theirs.

What if my budget is small?

Better to start this way. The first version is small on purpose: one problem only, with a defined goal. If it works, we grow it.

Do I need to have an IT team?

To validate the first version, no.To maintain and grow the product, it depends on the scope: maintenance can be carried out by me or by a development team.We define this together.

Does it work with what I already use, like ERP and spreadsheets?

It depends on what you use and how your systems allow access to data, for example via API.Before closing the scope, I understand what you already have and tell you clearly what you can call and what you can't.

Meus dados ficam seguros?

The data and infrastructure stay with your company.Everyone logs in, and each person only sees what they need.I use AI to make the system as safe as possible, but no system on the internet is 100% safe.If the project grows, security and infrastructure specialists come in.

Why not buy a ready-made tool?

Ready tool serves the medium process.If your process is what sets you apart, it's worth having something done for it.If one solves it, I'll tell you right from the start.

What if it doesn't work?

The goal of the first version is to understand what works.If no one uses it, you find out before investing time and resources, and then adjust or change the direction of the product.How much you can change depends on your business model, and we evaluate this together.

What problem do you want to see solved first?

Tell me where your team and your clients waste the most time.We'll see if it's worth building a first version with AI.

Agendar conversa de 30 minutos