About

About

I build practical systems across software, automation, data workflows, and applied AI.

This site is where I keep the work that is worth making public: project notes, technical write-ups, research threads, and the occasional essay about how modern work actually gets done.

My background sits in the overlap between software, data, automation, and operations. I like building systems that save time, reduce ambiguity, or make a messy process easier to repeat. Sometimes that means a small script. Sometimes it means a data pipeline, a local AI workflow, a public-data analysis, or a tool that turns a one-off task into something dependable.

I am less interested in looking busy than in showing real work: what I built, what I learned, what broke, what improved, and what I would do differently the next time.

If you are reviewing the portfolio, start with the Projects page. It gives a curated list of data science, applied machine learning, automation, and research work, with short summaries and reproducibility notes.

What I work on

Most of my attention goes toward tools and workflows that are useful under normal working conditions, not just in demos.

That usually puts me near:

  • Software that supports repeatable work
  • Automation for research, reporting, and operations
  • Data workflows, applied machine learning, and evaluation
  • Local AI tooling, agents, retrieval, and structured knowledge
  • Public-data research, especially where the data is useful but awkward
  • Remote work, contracts, employment systems, and the practical side of independent technical work

How I think about building

I tend to trust boring systems that produce useful output. A good tool should make the next step clearer. It should remove a little friction, leave a trail, and be understandable when I come back to it later.

In practice, that often means combining:

  • Small software tools
  • Local or self-hosted models when they make sense
  • Retrieval and structured knowledge
  • Workflow automation and reporting
  • Interfaces that make repeated work easier to do correctly

I am drawn to the unglamorous middle of projects: cleaning inputs, checking assumptions, making outputs reproducible, documenting the tradeoffs, and turning a fragile experiment into something that can survive real use. That is where a lot of the value is.

What to expect here

Most of what I publish here falls into a few buckets:

  • Project write-ups and build logs
  • Technical notes on tools, data, automation, and applied AI
  • Research based on public datasets and open sources
  • Notes on remote work, contracting, employment, and the economics around software work
  • Lessons from experiments that were useful, surprising, or just stubborn enough to document

Some posts will be polished essays. Others will be more like field notes. I am okay with both, as long as they are useful and honest.

The goal is for this site to feel current, grounded, and worth returning to.

Contact

© Justin Stone.