Software Engineer · Data Science · Machine Learning

Maria Fonseca

Software Engineer exploring Data Science & Machine Learning.
I build data-driven applications, explore patterns in data, and turn machine learning ideas into practical software.
View my projects
About

From shipping software to reading data.

I started as a software engineer building mobile applications, focused on shipping reliable, well-structured products. Over time, the questions I found most interesting moved upstream — not just how a feature works, but what the data behind it says.

That curiosity led me into data science and machine learning. I enjoy exploratory data analysis: digging through a dataset to find patterns, anomalies, and the small inconsistencies that usually matter most.

I'm not interested in models for their own sake. I care about end-to-end products — the full path from raw data to something a user can actually open and use.

Currently exploring
Data ScienceMachine LearningData Visualization Statistical AnalysisPythonSQL
Experience

Six years in, still shipping.

Software Engineer
2020 — Present
AndroidKotlinPython
Featured project

Finance Analytics

A personal analytics platform that turns a CSV export of bank transactions into spending trends, anomaly detection, recurring-subscription detection, and explainable insights — processed and stored entirely on-device, with every threshold documented and evidence-motivated rather than fitted to a model.

View on GitHub
01CSV Import
→
02Enrichment
→
03Anomaly Detection
→
04Recurring Detection
→
05Insights Engine
→
06FastAPI
→
07Android App
247
Automated tests
4
Analytics modules
3
Engineering layers
100%
Rule-based & explainable
Finance Analytics overview dashboard screenshot Finance Analytics insights screen screenshot Finance Analytics transactions screen screenshot
KotlinJetpack ComposeMVVM FastAPIPythonPandas Roompytest
Coming soon

What I'm building next.

Event Impact AnalyticsIn progress
An observational study measuring whether NYC Yankees home games shift taxi activity around the stadium — built on a full year of NYC TLC Yellow Taxi trip records, taxi zone geometry, and the Yankees' 2019 schedule, framed around association rather than causal claims until the evidence supports more. Currently in the data-acquisition and validation stage: schema, quality and coverage checks on the taxi dataset are underway before any modeling begins.
PythonDuckDBPyArrowpytest
Data science journey

How I got here.

Software Engineering
Python & Data Analysis
Exploratory Data Analysis
Machine Learning
End-to-End ML Applications
Skills

Tools & areas of focus.

Data Science
PythonPandasNumPyMatplotlibSQLEDAData CleaningData VisualizationFeature Engineering
Software Engineering
KotlinAndroidJetpack ComposeGitREST APIs
Currently learning
Machine LearningScikit-learnStatisticsModel EvaluationFastAPIMLOps fundamentals
GitHub

Explore the work behind the projects.

Code, experiments, analysis and the process behind my projects.

View GitHub profile
Contact

Let's connect.