Data and AI, built end to end.
I build the data pipelines and the AI on top of them, from raw source to the tool your team opens every day. Freelance and hands-on: we scope it together, then I build it, whether it stands alone or slots into what you already run.
↓ scroll to begin
I build data and AI systems, and I take the build on start to finish: the plumbing underneath and the tool on top of it.
I grew into the whole chain one role at a time. I started in QA test automation, then moved through data analysis, data and software engineering, and product ownership, each layer teaching me the one above it. That's exactly why I like the work: every build asks something new of me, and it gets my best, every time.
worked across
the substrate · underneath all three
the work, up close
I start from what you need and end at something that earns its keep. Three domains do the work, and they overlap far more than they line up. Here's what each one looks like up close.
data engineering
scattered, made usableMost data starts out scattered, and a lot of it you don't have yet. I build the pipelines that go and get it: scraping the web, calling external APIs to enrich it, tapping legacy databases, files, and live event streams. Then I clean and combine the lot into one queryable source you can build on.
machine learning · the lifecycle loop
the whole chain, in one loopI take a problem from raw question to running system: the data science to understand it, the modelling to solve it, and the MLOps to deploy, serve, and keep it healthy. The whole loop, owned by one person.
AI engineering
woven in, or standing aloneAI is the means, not the product. I build solutions that use it, picking the right technique for the problem at hand. Often that's a large language model; just as often it's classical ML, vision, or speech, and now and then a few of them working together. What I hand over is a solution that runs in production.
the craft, on one dial
Same craft across all three domains. The filled dots are work I've shipped; the hollow ones on the outer edge are what I'm growing into next. Hover any dot to read it, or a domain name to isolate that sector.
Have something
in mind? Let's talk.
freddy@lonnin.comFreelance, and at home in an agile team. For anything bigger than one person, I bring in a network I trust. Happy with longer engagements too, so tell me what you're after.