Matheus is available for hire
Hire MatheusMatheus Faustino
Verified Expert in Engineering
Data Engineer and Developer
Frankfurt, Hesse, Germany
Toptal member since May 11, 2026
Bio
Matheus has written software since 2011 and has worked on data platforms for the last six years. On a SAS to Databricks migration on Azure, he designed the testing and integration and helped set up Unity Catalog access control and DataHub lineage. At Thinksurance he rebuilt the data pipelines on Airflow and NiFi, and one back-office process went from 3 days to 1 hour. Day to day he uses Python, SQL and dbt, and he has led a team of 7 developers.
Portfolio
Time:Matters
Python, SAS, Unity Catalog, DataHub, SQL, Azure, RAG Pipelines, Automation...
CreditShelf AG
FastAPI, GitHub Actions, Kubernetes, Argo CD, Docker, Automation, Python...
Allianz Global Investors GmbH
CI/CD Pipelines, Azure, Kubernetes, GitHub Actions, JFrog, GitHub
Experience
- SQL - 14 years
- Docker - 9 years
- Python - 8 years
- Databricks - 3 years
- Apache Airflow - 3 years
- DataHub - 3 years
- Apache NiFi - 2 years
- Data Build Tool (dbt) - 1 year
Work Experience
Data Engineer
2023 - PRESENT
Time:Matters
- Architected the testing and integration processes for the SAS to Databricks migration of a global logistics data platform.
- Integrated and tested the new platform's data flow: Azure Data Factory loads into ADLS Gen2, then into Delta bronze, silver and gold layers with data quality checks at each layer. Raw loads can be replayed.
- Co-designed access control and data obfuscation in Unity Catalog for the new Databricks platform.
- Implemented DataHub lineage for the legacy data warehouse. The compliance team now uses it to approve data access requests.
- Built nightly PostgreSQL ingestion (full loads and SCD Type 2) and the Power BI data model with row-level security.
- Built a local RAG system with LangChain, Ollama and ChromaDB that searches documentation in SharePoint, Confluence, DataHub and internal servers. It combines BM25 and vector search.
- Automated several manual server maintenance tasks, so the team spent less time on upkeep and more on feature work.
Technologies: Python, SAS, Unity Catalog, DataHub, SQL, Azure, RAG Pipelines, Automation, Microsoft Power BI, Azure App Service, Data Pipelines, Data Engineering, ETL, Claude Code, AI Engineering, Data Analytics, Technical Leadership
Software Developer
2023 - 2023
CreditShelf AG
- Extended the team's existing FastAPI microservice templates to standardize secure deployments with GitHub Actions CI/CD.
- Joined the team's ongoing move from Ruby on Rails to FastAPI and migrated some of the services.
- Built Python microservices with FastAPI and ran their DevOps, deploying them on Kubernetes with ArgoCD together with the infrastructure team.
Technologies: FastAPI, GitHub Actions, Kubernetes, Argo CD, Docker, Automation, Python, Microservices
Senior DevOps Engineer Consultant
2023 - 2023
Allianz Global Investors GmbH
- Built new reusable GitHub Actions CI/CD workflows on top of the ones the team already had.
- Joined a project that added JFrog artifact and vulnerability scanning to the CI/CD workflows for Kubernetes deployments on Azure.
- Supported the company's teams with their Kubernetes deployments on Azure.
- Led an offshore engineering team that built more reusable workflows and rolled them out across the company.
Technologies: CI/CD Pipelines, Azure, Kubernetes, GitHub Actions, JFrog, GitHub
Data Developer
2020 - 2023
Thinksurance
- Replaced a back-office benchmark that a colleague ran by hand from scattered scripts, up to a week per request, with one Airflow pipeline. Processing went from 3 days to 1 hour.
- Redesigned the company's Airflow data pipelines on AWS to fail fast, with Presto data quality gates after every ingestion step that stopped downstream tasks before bad data reached reports.
- Built event ingestion from the company's microservices. Messages came in through Amazon SQS, and Apache NiFi checked them against Avro schemas, routed them, retried failures and wrote them to Hadoop.
- Built data marts on Hive for the analytics team and wrote Java UDFs for the heaviest transformations.
- Ran SQS and NiFi across several environments with alerts and load dashboards, and traced an incident where backed-up queues filled NiFi's memory and disk and stopped ingestion.
- Built a Selenium framework that simulated user profiles and triggered set event flows, producing realistic test data for checking warehouse ingestion.
- Helped move the data pipelines from the monolith to the company's new Kubernetes microservices.
Technologies: Python, Apache Airflow, Java, Kubernetes, Presto, SQL, Apache NiFi, Apache Hive, User-defined Functions (UDF), Data Pipelines, Selenium, ETL, Hadoop, Apache Avro, Data Modeling, Amazon Simple Queue Service (SQS)
Lead Developer
2017 - 2020
Dicas Mei
- Architected a data ingestion platform on AWS with Java (Spring), MongoDB and RabbitMQ that processed MEI seller data for clients, including major Brazilian banks and consumer goods companies.
- Led a team of 5 developers on the B2B data integration services, from design to delivery.
- Set up CI/CD pipelines in GitHub Actions that built the team's services and shipped them to AWS.
Technologies: Java, Spring, MongoDB, RabbitMQ, CI/CD Pipelines, Team Leadership, MySQL, PHP, Symfony
Full Stack Developer
2013 - 2016
Arca Solutions
- Worked on eDirectory, a white-label platform that let customers run their own directory sites with listings, events, classifieds and newsletters, and take payments through Stripe.
- With a team of 3 developers, moved eDirectory from an in-house PHP framework to Symfony 2.7 LTS and from PHP 5.1 to PHP 7 to cut technical debt and shorten release cycles.
- Brought in modern tooling and Elasticsearch search on the new stack, which the old framework could not support.
- Built features in PHP with Symfony and Twig, MySQL, jQuery and Bootstrap, including scheduled cron jobs and Memcached caching.
- Built features in PHP with Symfony and Twig, MySQL, jQuery and Bootstrap, including scheduled cron jobs and Memcached caching.
Technologies: PHP, Symfony, MySQL, jQuery, Bootstrap, Elasticsearch, Memcached, Docker, Twig
Web Developer
2011 - 2013
2RS Soluções em TI
- Developed a white-label e-commerce platform in PHP 5.3 with Zend Framework, which generated branded online stores for each client.
- Worked across the whole stack, from Zend Framework controllers and models to Bootstrap 2 and jQuery pages that had to work in Chrome, Firefox and Internet Explorer 6.
- Helped build a penny auction site, where each bid raised the price by 1 cent and the last bidder won the product when no one else bid.
Technologies: PHP, Zend Framework, PHP MVC, jQuery, Bootstrap, JavaScript, eCommerce
Project History
B2B Data Warehouse | dlt, dbt and Airflow ELT
I built an ELT data warehouse for a B2B e-commerce business, using synthetic data. dlt loads three sources: a PostgreSQL database, web server access logs enriched with device and country, and an Excel file of marketing leads. In dbt I modeled staging, dimension, fact and report tables, kept user and price history with snapshots, and tested every layer, with enforced contracts on the sales marts. Airflow runs the pipeline daily and tests each model right after it builds.
Enterprise Knowledge RAG | Local LLM Search
At time:matters I built a local search tool over the company's documentation in SharePoint, Confluence, DataHub and internal servers. It runs on local models, so no data leaves the company. Search mixes keyword and semantic matching, which finds exact table names as well as questions asked in plain language. Counting questions go to a restricted pandas layer instead of the model.
Startup Profiling Engine | Information Extraction and Data Mining
https://drive.google.com/file/d/1-QTMn6A11xaUpoYrcZh6QEKppWOz8rno/view?usp=sharingResearched and built an automated startup profiling system at the Grenoble Computer Science Laboratory (LIG) as part of a 21-week research internship at Grenoble INP. The system extracted and structured corporate data from company websites and third-party sources to generate investment-readiness profiles for informatics startups.
The engine combined 2 extraction strategies: a rule-based approach using XPath and Regex for structured data extraction, and an ML-based pipeline using binary classification and feature selection to identify which technology sectors each startup belonged to. Three classification algorithms were evaluated per sector to select the best-performing model, addressing the challenge of obtaining representative training samples across diverse tech domains.
The final system successfully populated over 50% of the proposed profile fields, surfacing previously uncollected corporate data points.
The engine combined 2 extraction strategies: a rule-based approach using XPath and Regex for structured data extraction, and an ML-based pipeline using binary classification and feature selection to identify which technology sectors each startup belonged to. Three classification algorithms were evaluated per sector to select the best-performing model, addressing the challenge of obtaining representative training samples across diverse tech domains.
The final system successfully populated over 50% of the proposed profile fields, surfacing previously uncollected corporate data points.
Stock Radar | Python Stock Signal Platform
Stock Radar is a Python tool I built that finds, scores and sends alerts on swing-trading setups using only free public data. A daily scanner ranks the watchlist by technical indicators, analyst data and trend, and stores the history with SQLAlchemy. It also watches SEC EDGAR for new filings and insider buys on tracked stocks. Alerts go to Discord, and it runs in Docker on my homelab with CI in GitHub Actions.
Education
2013 - 2018
Bachelor's Degree in Information Systems
São Paulo State University - Bauru, Brazil
2016 - 2017
Student Researcher Program (BRAFITEC Scholarship) in Information Systems Engineering
Grenoble Institute of Technology - Grenoble, France
Skills
Libraries/APIs
jQuery, SQLAlchemy, Pandas
Tools
Apache Airflow, DataHub, Apache NiFi, RabbitMQ, GitHub, Claude Code, Microsoft Power BI, Azure App Service, Apache Avro, Amazon Simple Queue Service (SQS), Pytest, dltHub, Okapi BM25
Languages
SQL, Python, PHP, XPath, Java, Regex, SAS, JavaScript
Frameworks
Symfony, Presto, Spring, Selenium, Twig, Hadoop, Bootstrap, Zend Framework, PHP MVC
Paradigms
Automation, ETL, Microservices
Platforms
Linux, Docker, Azure, Databricks, Kubernetes, Ollama
Storage
Data Pipelines, MySQL, MongoDB, Apache Hive, PostgreSQL, Elasticsearch, Memcached
Other
Software Engineering, Data Engineering, CI/CD Pipelines, Web Scraping, Data Mining, Unity Catalog, GitHub Actions, RAG Pipelines, User-defined Functions (UDF), Team Leadership, Data Analytics, Technical Leadership, ELT, Artificial Intelligence (AI), Machine Learning, Natural Language Processing (NLP), Feature Selection, Binary Classification, FastAPI, Argo CD, Text Classification, Computer Networking, Operating Systems, Information Security, IT Governance, JFrog, Discord Bots, EDGAR, LLM Integration, AI Engineering, Data Build Tool (dbt), Data Modeling, LangChain, ChromaDB, eCommerce
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