Mountains

Pedro Blaya Luz

Former biologist, found my vocation in data and AI. I've built end-to-end data solutions across multiple industries: dashboards and analysis, data pipelines, predictive models, and agentic AI.

AI/ML Solutions Architect

Data Scientist • Data Analyst • Software Engineer

GitHubLinkedIn

About Me

Pedro Luz

Pedro Luz

AI/ML Solutions Architect at CAST

Working with

Banco do Brasil

I started my career in data while working as a biologist. During my Master's in Ecology, I discovered my passion for statistics and programming analyzing sea turtle data. In 2020, I decided to dive fully into technology, and since then I've built a strong track record in data, AI, and software, leading end-to-end projects across companies in different industries.

My first tech role was in a real estate tech company, building ETLs, predictive models, and dashboards. Afterwards, in a subscription-based book company, I took on a hybrid role as tech lead and data product manager—while still coding. Today, in the financial sector, I work as an AI/ML solutions architect applying NLP techniques and building machine learning models to power a financial products recommendation system as well as deploying AI agents in the strict and secure environment of banking.

My passion for staying current with state-of-the-art techniques means my clients get solutions built with the latest and most effective approaches. I'm always interested in understanding business challenges and building solutions that create real impact.

Experience

Key Achievements

  • Design and deploy a multi-method Recommendation System that intelligently matches customers with suitable financial assets
  • Engineer AI agents in a secure banking environment, balancing compliance with operational efficiency

Technologies

PythonSQLPandasPySparkDatabricksFastAPILangChainDockerKubernetesOpenShift

Education

UFRGS

Master's degree, Ecology

Universidade Federal do Rio Grande do Sul

2017 – 2019

Loggerhead and green turtle population dynamics in the western South Atlantic: insights from stranding patterns

Data analysisStatistical modellingGeneralized linear models
R
UFRGS

Bachelor's degree, Biology

Universidade Federal do Rio Grande do Sul

2011 – 2015

Análise dos padrões de encalhes de tartarugas marinhas no litoral norte e médio do Rio Grande do Sul entre 1994 e 2015

Languages

🇧🇷

Portuguese

Native/Bilingual

Brasileiro nascido e criado. Viajei o país todo e virei sommelier de sotaques: sou apaixonado pelas mil maneiras diferentes de falar português.

🇺🇸

English

Fluent

English isn't something I learned. It's something I grew up with. Over twenty years of studying, working, and building friendships made it feel less like a second language and more like a second home.

🇪🇸

Spanish

Proficient

América Latina ha sido parte de mi camino. Uruguay y Argentina, en especial, ya no se sienten como viajes: en Uruguay tengo familia muy cercana y comparto el mate, la cultura y el corazón cisplatino con mis hermanos latinoamericanos.

Tools & Technologies

Here's a snapshot of some tools I work with day to day. But let's be clear: tools are interchangeable.

The real foundation is understanding how systems are conceived, built, and stressed in the real world. How ideas turn into architectures, and how those architectures evolve over time.

Frameworks and libraries get replaced. Principles endure. If you understand the fundamentals, picking up new tools is just an implementation detail.

Languages

PythonRSQL

Dashboard & Visualization

Power BILooker StudioTableau

Data & AI/ML Libraries

PySparkspaCyXGBoostPandasNumPyScikit-learnPyTorchTensorFlowLangChain

Python Development

FastAPIJupyterPydanticPytestPyPI

Cloud Platforms

AWSDatabricks

DevOps & Tools

GitGitHub ActionsDockerKubernetesOpenShift

Databases

PostgreSQLMongoDBOracleSQLiteFirebase

Bonus

Guide the turtle to the sea to reveal memories from my previous life as a biologist.

Sea turtle

Reach out

Happy to talk about projects, collaborations, or anything in between.

LinkedIn

© 2026 Pedro Blaya Luz. Built with Next.js, GSAP, and Tailwind CSS.