Software · AI · Data · São Paulo, Brazil
We build custom software, AI agents and data pipelines for companies that need to go from problem to product in weeks, not months.
What we do
01
From multi-tenant SaaS to internal dashboards. Installable web apps, mobile-ready, with the integrations that keep the business moving.
02
Claude-powered agents that do real work: they answer, classify, generate content and grade, with a human in the loop where it matters.
03
Pipelines that don't break at 3 a.m. Ingestion, modeling and dashboards on Google Cloud, with infrastructure as code.
How we work
We research the market, talk to the people who live the problem, and only then decide what to build.
Every decision becomes a document before it becomes code. You approve the design, not the surprise.
Automated tests, code review and small, frequent deliveries, so you watch the product grow.
Live on Google Cloud, with monitoring. And we stay with you after launch.
Projects
In production
An English learning platform with an AI teacher for every student.
The problem. English schools juggle six tools to manage students, classes, grades and payments.
The solution. One platform: CRM, scheduling, grades, payments and an AI teacher that grades pronunciation, generates lessons and remembers each student's weak spots.
Node.js · PostgreSQL · Claude · Cloud Run · Stripe · Asaas
MVP launching
Local auto parts finder: the right part, at the nearest store.
The problem. Mechanics and couriers lose hours calling store after store looking for a part.
The solution. Stores connect their Mercado Livre account, the app imports the catalog and shows who has the compatible part a few kilometers away.
Next.js · PostgreSQL · Mercado Livre API · Cloud Run
In pilot
Revenue recovery over WhatsApp.
The problem. A quote sent and forgotten becomes money lost in the middle of a chat.
The solution. AI reads the conversations, finds stalled opportunities, suggests the follow-up at the right time and shows how much came back.
Python · Claude · WhatsApp Business API
About
Data engineer with over six years of experience, focused on Google Cloud: BigQuery, Cloud Composer, Dataflow, Pub/Sub and Cloud Run. Also Databricks, Spark, AWS, Azure and Terraform. Databricks Certified Data Engineer Associate.
I like turning messy, manual processes into simple, repeatable systems. At ZAMP, I rebuilt the ingestion layer on Cloud Run and Terraform: adding a new data source went from a two-day task to a 30-second command, with lower infrastructure costs. At Stone, I fully automated a credit-acquisition process, from S3 into a Medallion architecture on Databricks and GCP. At DASA, I lead the migration of Azure Data Factory pipelines to Cloud Composer.
LV OLIVEIRA TEC grew out of the products I build after hours: ACE Learning, in production with real schools and payments; Bora Peça, integrated with Mercado Livre; and Cutucada, which uses AI to recover revenue on WhatsApp. It is the same method I bring to client projects: research before code, spec before screens, tests before deploy. I spent a year living in Canada and work with teams in Brazil and abroad, in Portuguese and English.
Contact
Tell us in a few lines what you need. We reply within one business day.