Taking on new projects
We build custom AI solutions, advanced automation and precisely engineered software — with ultra-fast delivery and the best tools on the market. A small team, direct contact with the people writing the code, zero middle layer.
The stack we work with
n8n
LangChain
Anthropic Claude
Next.js
React
TypeScript
Python
PostgreSQL
.NET / C#
WhatsApp Business API
Chrome Extensions
Vercel
Docker
Cloudflare
Tailwind CSSLebtag is a software development boutique specialized in the AI ecosystem. We work with the scope defined before we start: every project is designed, built and shipped by engineers who own the entire stack — from the AI agent to the desktop installer.
Our edge is speed with precision: we use the right tools instead of rebuilding everything from scratch, and we ship in weeks what a traditional agency ships in months.
Intelligent workflows connecting your systems, your data and language models. No-code/low-code automation, API and ERP integrations, conditional triggers, document processing and routines that run on their own 24/7.
See detailsAI agents that actually execute tasks, RAG systems (Retrieval-Augmented Generation) that answer from your knowledge base, advanced chatbots and custom tools wired into your business rules.
See detailsRobotic Process Automation and intelligent data extraction at scale: bots that operate portals with no API, collect and normalize information, and feed your systems with reliable data.
See detailsWeb, mobile, desktop applications and Google Chrome extensions. From a testable MVP to a product in production, with an architecture built to last.
See detailsTwo ways to work with us. The engagement model is assessed case by case, according to the project.
The default path. We understand the problem, map the complexity and define the scope with clear deliverables and deadlines from day one. The engagement model is assessed according to the project, once both sides know exactly what will be built.
For genuinely promising and innovative ideas, we step in as technology partners: we take on building the product and share both the risk and the outcome. If you have the idea and the market, we have the engineering.
We are a boutique: a small, senior, closed team. The person you talk to is the person writing the code — there is no management layer between the problem and the engineering.
We work by scope, with short, verifiable deliveries. Every project has one technical owner from start to finish, and what we ship goes to production with monitoring, documentation, and a client who knows how to operate it.
Nothing goes into construction before the problem is mapped and the expected outcome is written down. That protects both sides: you know what you are getting, and we know what we have to deliver.
We use the right tools instead of rebuilding everything from scratch. The first working deliverable usually ships within the first two weeks.
The model reads, extracts and converses; business decisions stay in explicit, testable rules. A system that decides on its own with no trace does not reach production.
Code, infrastructure and credentials are yours. We hand over with documentation and a proper handoff — we don't build vendor lock-in.
A few of the systems we built and keep running in production.
The data sits in the ERP, but the people who need it never open the ERP. A manager asks in a group chat, someone drops what they were doing, exports a spreadsheet and answers half an hour later.
Outcome: A lookup that used to depend on a person becomes a message answered in seconds, around the clock.
Before auditing anything, someone has to gather the invoices — and pulling XML files from the tax authority, company by company, already eats half a day. Then comes checking tax classification across thousands of line items by hand, repetitive work where the mistake only surfaces once it has become a penalty.
Outcome: Collection and checking stop being manual work: invoices arrive on their own and review becomes exception-based, with a trace of which rule flagged each discrepancy.
Brazilian rural land tax is an annual, low-ticket, high-volume filing. Firms were spending qualified hours on repetitive calculation and on hunting down land values municipality by municipality.
Outcome: Firms file in minutes what used to take qualified hours, with the land-value database built in and no retyping into the tax authority's software.
Analysis took a full business day, and nearly all of it was collection — opening seven different sources, taking screenshots, pasting them into the report. The analysis itself took minutes.
Outcome: From a full business day to minutes, with a PDF report, attached evidence, and every conclusion traceable to the rule behind it.
If yours isn't here, send it over on WhatsApp — we usually reply the same day.
The model is assessed according to the project. First we understand the problem and map the complexity; from there we define scope, deliverables and deadlines, and only then do we discuss the engagement format — which can be a defined project or a technology partnership, when the idea is promising enough for us to share the risk. To size anything we need to know what goes in and out of the process you want to automate, and that is what the free fifteen-minute call is for.
We work in weeks, not quarters. The first working deliverable usually ships within the first two weeks, because we use mature tools — n8n, LangChain, Next.js — instead of rebuilding everything from scratch. Larger projects are sliced into short, verifiable deliveries so you see something running well before the full scope is done.
Both, and the choice depends on the problem. n8n is better when the challenge is integrating systems and orchestrating a flow with predictable steps — every node is visual and auditable. LangChain comes in when language reasoning is required: agents that decide which tool to call, RAG systems answering from a knowledge base, chatbots wired into business rules. In practice most of our production systems combine the two.
Yes. Code, infrastructure and credentials are yours from start to finish. We hand over with documentation and a proper handoff, so your team can operate and evolve the system without depending on us. We don't build vendor lock-in.
Only if you want it to — and usually we advise against it. Our default pattern is that the model reads, extracts and converses, while business decisions stay in explicit, versioned, testable rules. In the credit pre-analysis engine we built, for instance, AI reads documents and images, but no credit conclusion is generated by a model. A system that decides on its own without leaving a trace does not reach production.
Yes. We are based in Goiás, Brazil, and work remotely with clients across Brazil and abroad, in Portuguese and English. The whole process — discovery, delivery, review and handoff — runs remotely, and that is how we already run the systems we keep in production today.
You describe the process you want to automate and you leave with a technical path: which tools to use, in what order, what is straightforward and where the risk sits. It is not a sales meeting: we don't present a commercial proposal in that call. If it makes sense to work together afterwards, then we talk about scope.
06 — Contact
Tell us what you need. We reply with a technical path and the next steps — no one-hour discovery call before that.
WhatsApp: +55 62 99381-5020·We reply during business hours, usually the same day.
Prefer email?lebtagdev@gmail.com