Ian Solberg

As a marketing analyst co-op at Wayfair, most of my work is AI development and n8n automation. The team's brief is internal tooling that just works: things colleagues use without being taught, and without asking anyone to run them. Alongside that I work in BigQuery, pulling and validating data and writing the cleaners and analysis scripts that make recurring questions repeatable, because establishing that a dataset can be trusted is most of the job before any conclusion is drawn from it.

Alongside that I build research tooling and applications for organizations without dedicated software teams. The through line is data access: a research group whose measurements could not be explored without writing code, a farm whose records were held in a single Access database file, a school's faculty who could not see their own budget lines. Each deliverable was an interface built for the person who needed the data.

My open-source work follows the same pattern. fred-loader and census-loader provide readable naming layers over federal data APIs, with discovery helpers so a series can be found interactively rather than looked up in documentation.

Background

I grew up in New Hampshire, living in Bridgewater, Holderness, and most recently Hanover. In high school I ran track and played football, sang in the show choir, and led the male a cappella group. In college I joined the Northeastern track team as a high jumper and helped contribute as a scorer to two conference championships.

Interests

Most of the projects I take on are close to the organizations that use them, which is how the farm and school work started. The economics half of my degree is focused on distributional macroeconomics: how wealth concentration affects consumption, investment, and the transmission of monetary policy. My personal research tooling was built to examine those questions against real data.

I write about that at more length on my writing page and on Substack, most recently on how government debt above 100% of GDP changes the relationship between money supply and inflation.

Analysis: Python, pandas, SQL, BigQuery, JupyterLab, pandera for schema and dtype validation, NumPy.

Engineering: FastAPI, Flask, Django, SQLAlchemy, MySQL, Postgres, HTMX, Tailwind, Docker.

Practices: strict typing and linting, explicit validation over silent failure, documentation written for the next person to use the code.

Email is the best way to reach me.