Cassiano Schneider
AI Engineer, Santa Maria, Brazil
Eleven years across engineering and product, with multi-agent systems running in production on LangGraph, Google ADK, MCP and RAG.
- Python
- Django
- FastAPI
- LangGraph
- Google ADK
- MCP
- PostgreSQL
Who does the work, and how
The people who write the code, the technologies they use, and how the work runs from first contact to delivery.
Why Zorgen exists
Zorgen builds software for companies whose operation has outgrown spreadsheets, manual handoffs and tools that do not talk to each other. We are a small senior team: the people who scope the work are the people who write the code and answer when it breaks.
Who does the work
We are two senior engineers, in two countries.
Jorge Saicoski Sarkis
Backend & Cloud Engineer, Montevideo, Uruguay
Eight years in backend and cloud; at The Home Depot, via TCS, took Go API queries from ten seconds down to about two.
- Go
- Node.js
- AWS
- GCP
- Terraform
- Kubernetes
- Docker
Working together
Time zone
Brazil and Uruguay, both UTC−3. Full overlap with the US Eastern morning and afternoon.
Communication
Slack, or whichever channel your team already uses, with a written update every two days and a weekly call.
Language
English, Portuguese and Spanish.
Ways to work with us
Two arrangements, depending on what your company already has. Starting with one and moving to the other is possible, and we sort that out in conversation.
Inside your team
For teams that have the people and need more hands. We join the channel your team already uses, Slack or Teams, and your repository, whether that is GitHub, GitLab or Gitea. The process stays yours and we fit into it.
Work is measured per task, with the size agreed before the task starts. You know what each item costs before you approve it.
By project
For companies that have the problem and no team to build it. The team is ours and the scope is agreed with you. The first delivery is live in two weeks, and a new one every two weeks after that.
How the work runs
- Discovery: we learn how the operation actually runs before proposing architecture, because the written requirement is rarely the problem.
- Two-week cycles, with something running at the end of each one, instead of a single delivery at the finish line.
- Every change goes through code review and automated tests before it ships.
- We hand over with automated deploys and written instructions for operating what was built.
After delivery
Code ownership
The code is yours from the first commit. The repository lives in your organisation; we work in it by invitation.
Support
Thirty days of fixes at no cost after delivery, and a maintenance agreement when it makes sense.
Built to fit, not to fill a template
We dig into the problem until we understand it better than the brief did. That is what lets us shape the solution around how a particular company actually works, instead of around what was easy to specify.
What we stand for
Every engagement is judged by one thing: whether the client's operation actually got better.
Straight talk
We say what needs saying, including when a request is the wrong solution to the real problem.
Partnership
We measure our work by our clients' results, not by hours billed. That is why engagements here tend to last years, not sprints.
Support
We monitor what we ship and answer for it in production, not only until handover.
Team
Senior engineers who stay on your project. No handing the work down to whoever is free.
Modularity
Systems built so the next feature is an addition, not a rewrite.
Security and privacy
Data handled with care and in line with Brazil's LGPD, the local equivalent of the GDPR.
Integration
We connect to the systems you already run (ERPs, messaging platforms, payment providers) instead of asking you to replace them.