São Paulo, Brazil · remote

Hi, I'm Bruno Sartori.
Senior software engineer.

I'm a backend developer from Brazil, working mostly with C# and .NET these days. Before that I spent years on the front-end: React, TypeScript, multi-tenant platforms and OTT apps for Samsung Tizen, LG webOS and Chromecast. Lately I've been going deep on AI. Agents help me research, plan, write and test code every day, and the calls on what ships are still mine.

years shipping software
10+years shipping software
backend core, C# first
.NETbackend core, C# first
Tizen · webOS · Chromecast
OTTTizen · webOS · Chromecast
human-in-the-loop by default
HITLhuman-in-the-loop by default

agent://session/billing-api

live

mode plan → actapprovals requiredsession 1/3

↑ simulated sessions, modeled on how I actually work

01Human × Machine

AI is my pair, not my autopilot.

I treat agents like fast, tireless collaborators with clear boundaries. Every task runs through the same loop, and the steps where judgment matters stay with a human.

  1. 01◇◆ shared

    Context

    Curated knowledge bases, docs and ADRs, so agents start from facts instead of guesses.

  2. 02◆ human

    Plan

    The agent drafts a plan. I push back on it, cut scope and pick the trade-offs.

  3. 03◇ agent

    Execute

    Agents write code, run tests and reach real systems through MCP tools and custom skills.

  4. 04◆ human

    Verify

    Tests, diffs and review. Nothing is committed, deployed or published until a human approves it.

Toolkit

  • Agentic coding

    Claude Code is my daily pair programmer. I use plan mode for bigger changes, subagents for research in parallel, and hooks to enforce project rules.

    Claude CodeSubagentsHooks
  • Custom skills

    Packaged, reusable instructions that teach agents a team's workflows, such as syncing a docs site into a project wiki from start to finish.

    SkillsWorkflows
  • Living knowledge bases

    Raw docs go in and a structured wiki comes out. An agent tracks release history, breaking changes and contradictions, all checked against the source.

    Context engineeringWikis
  • MCP integrations

    Connecting models to real systems through the Model Context Protocol: repos, trackers, APIs, even my anime list.

    MCPTool use
  • Personal agents

    Small autonomous helpers that search and rank things against explicit criteria. The final call stays with me.

    AutomationRanking
  • Guardrails first

    Permissions, approvals and review gates are built into the loop. Agents move fast, and anything irreversible waits for a human.

    HITLSafety

02Stack

A backend core, a front-end past, an AI‑native present.

I started on multi-tenant web platforms and OTT apps for Samsung Tizen, LG webOS and Chromecast, with video controls, analytics and SEO along the way. Today my center of gravity is backend engineering in C#, with AI agents built into how I deliver.

Engineering core

  • C#
  • .NET
  • TypeScript
  • Node.js
  • React
  • Next.js
  • React Native
  • AWS
  • Samsung Tizen
  • LG webOS
  • Chromecast
  • OTT & video
  • Analytics & tagging
  • SEO

AI toolkit

  • Claude Code
  • Claude API
  • Model Context Protocol
  • Agent skills
  • Subagents
  • Hooks
  • Context engineering
  • Knowledge bases
  • Prompt design

Education

  • Postgraduate degree

    Internet of Things

  • Degree

    Systems Analysis & Development · IFSP

off-duty: anime · Bujinkan ninjutsu · sci-fi · philosophy · quantum computing

05 · Contact

Open a channel.

Have a role, a project or a hard problem in mind? Want to talk agents, OTT or anime? My inbox is open.