Data Scientist · [City, State]
I build models that hold up outside the notebook — forecasting, causal inference, and the pipelines that keep them running in production. Currently looking for a full-time data science role.
About
I'm a data scientist with a background in [your degree] from [your school]. Most of my work sits where messy real-world data meets a decision someone actually has to make — which is usually less about the model and more about whether the question was framed right in the first place.
Before this I [one line of relevant background — a research role, an internship, a career switch, whatever gives context]. That's where I learned to care about reproducibility, because I've been the person who had to re-run someone else's analysis six months later.
Outside of work I [one genuinely human detail — climbing, bread, chess, birding]. It makes the interview conversation easier, and it's true.
Projects
Four projects that show the range: a forecast, a causal question, an NLP pipeline, and something built for other people to use.
Hierarchical time-series forecast across 1,200 SKUs and 40 locations, with reconciliation so store-level predictions actually sum to the regional plan.
Cut MAPE from 18% to 11% against the incumbent seasonal-naive baseline.
Difference-in-differences with a synthetic control on staggered rollout data, plus sensitivity analysis for how much unobserved confounding it would take to flip the sign.
Effect held at +4.2% retention under a 1.8x confounding stress test.
Fine-tuned a transformer to route incoming tickets to the right team, with a confidence threshold that hands ambiguous cases to a human instead of guessing.
91% routing accuracy; 34% of manual triage time recovered.
A public dashboard on city housing permits — the whole pipeline from scraping to a hosted app, rebuilt nightly and designed for people who don't read charts for a living.
~2k monthly users; cited by two local news outlets.
Experience & education
Company Name
One or two sentences on what you own and what changed because of it. Lead with the outcome, not the tool list — the tools are already on your resume.
Company Name
What you shipped in twelve weeks, and whether anyone still uses it.
University Name
Relevant coursework, honors, thesis, or a research group worth naming. Skip the GPA unless it helps you.
Contact
Open to full-time data science roles. If you're hiring — or just want to talk about a project — the fastest way to reach me is email.