Applied AI
Bounded tools, automation, evaluation, and useful workflows where model output must remain grounded in evidence.
Applied AI · Adtech · Real business problems
I use this site to build in public, examine how technology works in practice, and document what I learn. The scope is applied AI across real business problems, with deeper context from the adtech industry.
Projects appear at their real stage. Each one records the problem, the technical and commercial choices, the evidence, and what remains unresolved.
A personal project exploring how an AI assistant can ask for narrow evidence while deterministic code calculates the numbers.
The output is only one part of the record. I also document the choices behind it: what was included, what failed, what changed, and what the evidence can support.
Separate the business decision from the first technical request.
Make inputs, calculations, responsibilities, and limits explicit.
Check whether the result answers the problem, then state what remains uncertain.
Bounded tools, automation, evaluation, and useful workflows where model output must remain grounded in evidence.
Campaign economics, measurement, integrations, delivery, and the systems behind performance advertising.
How requirements, incentives, commercial constraints, and technical trade-offs shape real solutions.
About the lab
My career has largely been customer-facing. I have worked in technology since 2021, with recent experience in enterprise integrations and performance advertising.
This site is where I store projects and field notes as I grow technically and commercially. The portfolio is the accumulated record of that work.
Read about the lab →Follow or compare notes