Pedro Fernandes Steimbruch, Developer in Jaguarão - State of Rio Grande do Sul, Brazil
Pedro is available for hire
Hire Pedro

Pedro Fernandes Steimbruch

Software Developer

Jaguarão - State of Rio Grande do Sul, Brazil

Toptal member since December 19, 2016

Bio

Pedro is a senior full-stack engineer with two decades of experience building production SaaS from zero to scale. He specializes in Ruby on Rails and React, with recent focus on shipping LLM-powered products: RAG pipelines, multi-agent systems, and AI content automation. He has a track record of carrying products from prototype to $10M+ ARR with very small teams.

Portfolio

GrowthX
Ruby on Rails (RoR), React, Neo4j, TimescaleDB, DuckDB, PostgreSQL...
Cleary
Ruby on Rails (RoR), React, Amazon OpenSearch, PostgreSQL, LLM Integration...
CollateralXP
Ruby on Rails (RoR), Tailwind CSS, Stripe, PostgreSQL, Alpine.js, Turbo, CSS...

Experience

  • Ruby - 18 years
  • Ruby on Rails (RoR) - 18 years
  • JavaScript - 8 years
  • PHP - 8 years
  • CakePHP - 6 years
  • jQuery - 6 years
  • AngularJS - 3 years

Preferred Environment

Git, Claude Code, Cursor AI, Ruby on Rails (RoR), TypeScript, Vercel, Ruby, Full-stack Development, API Integration

The most amazing...

...work I've done is architecting ProxyPics from scratch, scaling it past $10M ARR with only four engineers, powering 10,000+ home inspections per month.

Work Experience

Senior Full-stack Engineer

2025 - PRESENT
GrowthX
  • Built a Ruby implementation of LightRAG — a knowledge-graph + vector RAG system on Neo4j with per-tenant isolation — serving as the research-phase knowledge base for AI content-generation agents.
  • Replaced manual GA/GSC reporting with a unified analytics dashboard (Rails + Inertia/React) backed by a 500GB+ TimescaleDB store with billions of rows, giving the team a single source of truth for content performance.
  • Designed end-to-end AI content pipelines (research → draft → fact-checking → internal linking → multi-CMS publishing across CraftCMS, Webflow, Sanity, WordPress, and others), shipping hundreds of articles per week.
  • Contributed to checkthat.ai — an Answer Engine Optimization index tracking brand visibility across Google AI Overview, Claude, ChatGPT, and Perplexity — using DuckDB over Parquet with a Postgres materialization layer.
Technologies: Ruby on Rails (RoR), React, Neo4j, TimescaleDB, DuckDB, PostgreSQL, LLM Integration, Amazon OpenSearch, Inertia, Python, CSS, Ruby, Amazon Web Services (AWS), AI-generated Code, Large Language Models (LLMs), Next.js, Vector Databases, Full-stack Development, API Integration, Artificial Intelligence (AI), Platforms, Third-party APIs, Agentic AI

Senior Full-stack Engineer

2023 - 2025
Cleary
  • Architected lazy-loading and rendering improvements to Cleary Pages (Rails API + React SPA), reducing load time by 4x.
  • Built a RAG pipeline blending vector DB and OpenSearch with user-level content authorization, serving as the foundation for Cleary's AI features.
  • Shipped an LLM employee-support agent serving employees across in-app, Slack, and email channels, and designed a supervisor-orchestrated multi-agent system over the RAG pipeline.
  • Co-designed a Zapier-style HR workflow automation built on AI triggers and function calling, automating HR ticket resolution end-to-end.
Technologies: Ruby on Rails (RoR), React, Amazon OpenSearch, PostgreSQL, LLM Integration, Slack, TypeScript, CSS, Ruby, Amazon Web Services (AWS), AI-generated Code, Large Language Models (LLMs), Vector Databases, Pgvector, Full-stack Development, API Integration, Artificial Intelligence (AI), Platforms, Third-party APIs, Agentic AI

Co-founder & Tech Lead

2021 - 2023
CollateralXP
  • Introduced ShapeUp-style product development across a team of five engineers, one PM, and one designer, improving delivery predictability and focus.
  • Architected a multi-tenant, async event-driven Rails modular monolith and shipped a rich customer-facing web app on Tailwind, Turbo, and Alpine.js.
  • Built an integration layer for multiple Loan Origination Systems, cutting time-to-onboard a new integration to days, and built a marketplace platform on Stripe Connect custom accounts.
Technologies: Ruby on Rails (RoR), Tailwind CSS, Stripe, PostgreSQL, Alpine.js, Turbo, CSS, Ruby, Amazon Web Services (AWS), TypeScript, Next.js, Node.js, Full-stack Development, React, API Integration, Platforms, Third-party APIs

Founding Engineer & Tech Lead

2018 - 2023
ProxyPics
  • Architected a multi-tenant, async event-driven Rails API and React/React Native front end that scaled the company past $10M ARR with only four engineers, supporting 10,000+ home inspections per month.
  • Led product design of the mobile app alongside one designer and built a Metabase-driven analytics practice that became the foundation for company strategy and roadmap decisions.
  • Shipped a public REST API serving multiple enterprise customers and third-party integrations, and built a Next.js gateway enabling new enterprise integrations to go live in days.
Technologies: Ruby on Rails (RoR), React, React Native, Next.js, Stripe, PostgreSQL, REST APIs, Metabase, CSS, Ruby, Amazon Web Services (AWS), TypeScript, Node.js, Full-stack Development, API Integration, Platforms, Third-party APIs

First Engineer

2016 - 2018
Gyve
  • Joined as the first engineer at a faith-based giving platform, building the core Rails API and React front end from scratch to support donors, churches, and nonprofits at scale.
  • Integrated Stripe for donation processing and built recurring giving workflows, enabling the platform to handle thousands of transactions monthly with a reliable payment infrastructure.
  • Designed and shipped background job processing with Sidekiq and Redis for async tasks such as email notifications, payment retries, and reporting, improving platform reliability.
Technologies: Ruby on Rails (RoR), React, PostgreSQL, Stripe, Redis, Sidekiq, CSS, Ruby, Amazon Web Services (AWS), Full-stack Development, API Integration, Platforms, Third-party APIs

Experience

SportSnap

https://sportsnap.io
SportSnap is a vertically integrated sports video platform I built solo and run as a profitable side business — in production with 6 paying customers.

The product turns a single physical button press at courtside into an instant video clip and live-streams entire games to YouTube.

I built every layer: ESP32 button-controller firmware in C++, an on-premise Python gateway that maintains rolling HLS buffers from PoE cameras and remuxes clips on demand, a Ruby on Rails cloud platform with a React 19 / Inertia admin UI on top of AWS (Lambda transcoding, ECS/Fargate live-streaming, Cloudflare R2 storage), and a player-facing app/social media to claim replays to profile. >1000 MAU.

Engineering wins: a serverless AWS Lambda + FFmpeg transcoding pipeline behind a pluggable interface, cutting per-video costs by 99%; an S3-to-R2 storage migration that cut storage costs another 99% by eliminating egress; and an edge gateway running on low-power RISC-V hardware with no port forwarding, no cloud relay, and no inbound firewall holes. The product spans firmware, embedded networking, video infrastructure, and SaaS — and I shipped all of it.

Education

2012 - 2015

Certificate in System Analysis

Senac Pelotas - Pelotas, Brazil

Skills

Libraries/APIs

jQuery, REST APIs, Google Maps API, Google Maps JavaScript API 3, Twilio API, React, Node.js, Fabric, Stripe, Sidekiq, FFmpeg, Inertia

Tools

Git, GitHub, Claude Code, Vim Text Editor, Subversion (SVN), Microsoft PowerPoint, Oracle E-Business Suite (EBS), Amazon ElastiCache, Sublime Text 3, Capistrano, Gulp, Amazon OpenSearch, Slack, Turbo, Amazon Elastic Container Service (ECS)

Languages

HTML, CSS, PHP, Ruby, JavaScript, TypeScript, CSS3, HTML5, Python

Frameworks

CakePHP, Ruby on Rails (RoR), AngularJS, Symfony, Materialize, Bootstrap, Prototype Framework, Symfony 2, Sinatra, Tailwind CSS, Alpine.js, React Native, Next.js

Paradigms

Object-oriented Design (OOD), Agile, Test-driven Development (TDD), Management

Platforms

Linux, Amazon Web Services (AWS), Amazon EC2, Heroku, AWS Elastic Beanstalk, AWS Lambda, Vercel

Storage

Redis, Memcached, PostgreSQL, MySQL, Neo4j, Amazon S3 (AWS S3)

Other

Full-stack Development, API Integration, Third-party APIs, Cursor AI, AI-generated Code, Large Language Models (LLMs), Vector Databases, Pgvector, Artificial Intelligence (AI), Platforms, Agentic AI, GNU, TimescaleDB, DuckDB, LLM Integration, Metabase, Full-stack, Planning

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring