[EXP] Experience

Machine learning that has to work outside a notebook.

I build healthcare ML pipelines at CRO Prime. Before that I did two internships covering customer churn, COVID-19 forecasting, text and image work.

Feb 2025 - Present

AI Engineer, CRO Prime

The problem: Healthcare teams hold medical data in many forms, structured records and free text, and turning it into predictions people can review needs a pipeline that gives the same result every run.
What I built: I built four-stage pipelines (ingestion, preprocessing, training and inference) that can be re-run reliably. I designed the medical data models so a prediction service can plug in, keeping accuracy and response time in mind, and I turned requirements from three stakeholder workflows into reusable ML components and review dashboards.
What came out: Delivery across those three stakeholder workflows got faster, because they share the same components and dashboards.

Internship

Data Analysis / Machine Learning Internship, remote.

The problem: A business wants to know which customers are likely to leave, early enough to act.
What I built: I framed it as a retention problem, prepared the structured data, chose the features that predict churn, and turned the model's output into a signal a team could act on.
What came out: A churn prediction used as a decision signal, and my first hands-on experience of ML for a business decision. I have no figures to share from it.

Internship

Data Analysis / Machine Learning Internship, remote.

The problem: Different kinds of data need different approaches: numbers over time, text and images.
What I built: I worked on three projects: COVID-19 case forecasting (time series), text pipelines for NLP, and image recognition for computer vision.
What came out: One internship that covered three areas of applied ML.