Skip to content
en
Book a call
Services

Three lines, the same engineering standard.

Weft builds (the first two) and guides (the third). You can hire one or combine them.

Builds

Software Engineering

Custom systems and products.

For those who need a system that doesn't exist yet, or an existing one working better.

  • SYSCustom systems: management, dashboards and automation for what lives in spreadsheets today.
  • MVPSaaS and MVP: from the idea to the first version used by real customers (an MVP is the minimum version to test with paying users).
  • WEBWebsites and landing pages: fast, findable on Google, on your own domain.
  • APIAPIs and integrations: systems that don't talk to each other start exchanging data on their own.
  • EVOLMaintenance and evolution: fixes, improvements and support after launch.
How a project runs
Builds

AI Engineering

AI working in operations.

For those who want AI doing real work every day, not just in slide decks.

  • AGENTAI agents: carry out a task end to end (triage, reports, customer service), with human review where needed.
  • BASEAssistant over the company knowledge base: answers from your documents and shows the source of each answer.
  • AUTOAI automation: the repetitive step leaves the team's hands.
  • SKILLSkills for AI tools: your company's way of working inside the Claude, ChatGPT or Copilot the team already uses.
  • EMBEDAI in the system you already have: without replacing what works.
How a project runs
Guides

Consulting

Assessment and team training.

For leaders who need to know where technology produces results, and for teams that want to build in their own context.

  • DIAGTechnical and AI assessment: prioritized opportunities and a 90-day plan.
  • TRACKMentoring tracks: an agent or solution in real use, built together with your team.
  • CLASSIn-company workshop: a group of up to 10 people, on your team's workflow.
  • MONTHMonthly support: sessions and support so the team can carry on alone.
See formats and pricing
How a project runs

Prototype approved before code. Scope agreed in writing.

Four gates, and nobody moves forward on assumptions, me included. Projects up to R$ 3,000 go straight from the briefing to the build.

G1Contract signed. No design or code work before this point, not even for free.
G2Prototype approved. Clickable, tested with the people who will use it. After it, any change becomes an amendment with its own price and deadline.
G3Acceptance testing done. You use it before you accept it.
G4Production sign-off. The warranty starts: 30 days on smaller projects, 90 on larger ones.

5-business-day approval SLA (the agreed response time) for each deliverable, on both sides. 2 projects at a time, at most.

Pricing

The price is right here, not behind a form.

Scope always agreed in writing: every change becomes an amendment with its own price and deadline. Prices in Brazilian reais (BRL).

Starting point

AI Assessment

R$ 2,5004H OF SESSIONS · 1 TO 2 WEEKS
  • Report with prioritized opportunities
  • Tool recommendations
  • 90-day plan
  • Credited toward a track if hired within 30 days
Start here
Recommended

Applied Track

R$ 8,500R$ 9,000 WITHOUT THE ASSESSMENT FIRST20H · 8 TO 10 WEEKS
  • One agent or workflow in real use
  • Documented so the team can replicate it
  • 2h per week + asynchronous review
  • Half of the Assessment fee credited
Book a call
Team transformation

Advanced Track

R$ 22,000R$ 24,000 WITHOUT THE ASSESSMENT FIRST40H · 12 TO 16 WEEKS
  • Agent architecture and multi-agent flows
  • Applied to the company's real projects
  • Patterns so the team can replicate on its own
  • Full Assessment fee credited
Talk about the team

In-company workshop

from R$ 12,0008H TO 16H · UP TO 10 PEOPLE
  • Immersion 8h · Program 12h · Full 16h
  • Content built around your team's workflow
  • From the 11th person on, R$ 800 per person
Talk to leadership

Monthly support

R$ 3,000/month4H/MONTH · 3 MONTH MINIMUM
  • 1 session per month + asynchronous support
  • For the team or for the delivered system
  • Hours don't carry over between months
Book a call
Custom software and AI
TrackRangeExamplesWhat you getWarranty
Expressup to R$ 3,000Website, landing page, change to an existing systemBriefing and delivery checklist30 days
StandardR$ 4,000 to 15,000Simple system, dashboard, integration, AI agentProduct Brief, approved prototype, handover30 days
Fullabove R$ 15,000Custom system, platform, MVP, multiple agentsEverything in Standard + technical spec90 days

Technical assessment of an existing system: R$ 3,500 to 7,000. The first 30-minute call is free; the discovery work (Assessment or Discovery) is paid, because it takes the hours that should be producing your result. Prices in Brazilian reais (BRL).

Questions

What people usually ask me on the first call.

Who does the work?

I do. Weft is a one-person operation, by choice: it is what allows quick answers, decisions without a committee and lower prices than an agency. The limit of that is written on the page: two projects at a time. When a project needs more hands, I say so before signing, not halfway through.

What if I don't hire anything after the assessment?

You leave with the report, the prioritized opportunities and the 90-day plan, and you can carry it out with anyone, including your own team. The assessment is paid precisely so that it is useful on its own. If you hire a track within 30 days, the Assessment is credited: 50% on the Applied Track and in full on the Advanced Track.

How soon will I see results?

The assessment is delivered in one to two weeks. An Applied Track puts an agent into real use in eight to ten weeks. A Build depends on the scope, but the clickable prototype is ready before any production code. So you see the product working long before it exists.

How much does a software or AI project cost?

It depends on the size: up to R$ 3,000 for a website or a change, R$ 4,000 to 15,000 for a focused system or agent, above that for platforms and custom systems. Values are in Brazilian reais (BRL). On the 30-minute call you already leave with the range for your case.

Do you use AI to write the code?

I do, every day, and it is part of why the price works. But AI speeds up engineering. It doesn't replace judgment. Architecture, scope decisions and reviews are still mine, and they are recorded in documents so you don't depend on me to understand your own system.

Do you work outside São José do Rio Preto?

Yes. All the work is remote, with video meetings and deliveries you follow in the browser. The base is in Rio Preto, in Brazil.

A 30-minute call answers the first question.

You bring the process, I tell you whether software or AI moves the number. And if it doesn't, I tell you that too.

Book a call
Book a callServices