Software Engineer

Murilo
Matos

Computer Science student at the University of Florida, building backend services, machine learning models, and the data pipelines that feed them. Most recently a software engineering intern at BTG Pactual in São Paulo.

Portrait of Murilo Matos

About

Systems that hold up
after you ship them.

I'm a Computer Science student at the University of Florida who likes the unglamorous half of software: the validation, the monitoring, the tests that catch a bad release before a client does. A reporting tool I built for an agriculture client has run in production for over 18 months and more than 600 reports without me touching it.

My work spans backend services in C#, Go, and Python; computer vision research at UF's Blueberry Breeding Lab; and data pipelines that turn messy sources into something a person can actually read. I'm bilingual in English and Portuguese, and I studied abroad at ZHAW in Switzerland.


Experience

Where I've worked.

  1. Software Engineering Intern

    Jun–Aug 2026

    BTG Pactual · São Paulo, Brazil

    Built a production backend service in C#/.NET for an internal platform managing the firm's financial positions, using MongoDB for persistence and RabbitMQ for asynchronous messaging. Designed a dynamic factory interface that lets new position types be added without modifying existing code.

    • C#/.NET
    • MongoDB
    • RabbitMQ
  2. Student Developer

    Jan 2026 – Present

    The Agency · University of Florida

    Ship full-stack features within an Agile team — Go backend, RESTful APIs, React front-end, SQL on AWS — maintaining reliability through automated testing, debugging, and code review.

    • Go
    • React
    • REST
    • AWS
  3. Machine Learning Research Assistant

    2024 – Present

    Blueberry Breeding Lab · University of Florida

    Two years of applied ML research: building computer vision models, curating datasets, designing evaluation methods, and communicating findings to collaborators across technical and non-technical backgrounds.

    • PyTorch
    • OpenCV
    • Research

Projects

Things I've built.

Hackathon · 2025

Clinix — AI Agent with RAG

A full-stack AI agent integrating an LLM with schema-validated tool-calling (Pydantic) to run multi-step workflows against external tools and data. Implemented retrieval-augmented generation over text embeddings and a vector database, plus guardrails and retry/backoff handling. Designed an automated evaluation harness that measures output accuracy against defined thresholds using smoke tests and API-contract checks, catching failures before release. Built and deployed by a 3-person team.

  • Python
  • FastAPI
  • React
  • OpenAI API
  • RAG

Freelance · Apr 2024 – Oct 2025

Fertigation Report Automation

I identified an operational bottleneck for an agriculture client and designed, built, and shipped a Python automation tool with data validation, cutting a manual reporting process from about an hour to under five minutes. Owned the system end-to-end as sole developer, sustaining 18+ months of production reliability across 600+ reports a year through monitoring, defect tracking, and documentation written for a non-technical client.

  • Python
  • Data validation
  • Automation

Research · 2024 – Present

Blueberry Crunchiness Classifier

A binary image classification model distinguishing crunchy from non-crunchy blueberries, achieving 90%+ accuracy. Curated and labeled a dataset of 3,000+ image records, trained convolutional neural networks, and benchmarked results against baselines with failure-case analysis. Part of ongoing machine learning research at the Blueberry Breeding Lab.

  • Python
  • PyTorch
  • OpenCV
  • CNNs

Research · 2024

Plant Phenotyping Analytics Dashboard

Data-analysis and dashboard tooling that transforms raw experimental measurements into structured summaries and visualizations, surfacing trends and metrics to support research decisions.

  • Python
  • Data viz
  • Dashboards

Course Project · 2026

Scrabble Word Finder — Trie vs. Min-Heap

Designed and implemented two data-structure approaches to word-finding, then quantified their trade-offs through time and space complexity analysis on large inputs to select the optimal approach. Built with a 3-person team.

  • C++
  • Tries
  • Heaps
  • Benchmarking

Skills

What I work with.

Languages

  • Python
  • C#
  • Go
  • Java
  • C++
  • JavaScript/TypeScript
  • SQL
  • Bash

AI & Machine Learning

  • PyTorch
  • scikit-learn
  • OpenCV
  • LLMs (OpenAI API)
  • RAG
  • Prompt engineering
  • Agents & tool-calling
  • Evaluation harnesses

Backend & Web

  • RESTful APIs
  • FastAPI
  • React
  • Next.js
  • RabbitMQ
  • .NET

Data & Databases

  • PostgreSQL
  • MySQL
  • MongoDB
  • Data pipelines
  • Data validation
  • Dashboards

Cloud & DevOps

  • AWS (S3, EC2, RDS)
  • Docker
  • Kubernetes
  • Linux
  • CI/CD
  • Git/GitHub

Foundations

  • Data structures & algorithms
  • Computer architecture
  • Object-oriented design
  • Design patterns
  • Complexity analysis

Contact

Let's build
something solid.