Working to make AI boring (predictable, reliable, safe). So we can use it to fix problems. Not to hype up AI.

I care about what works in practice and what we can use to solve problems. Not what could work or what is the state-of-the-art on a irrelevant benchmark.

I specialize in building complex AI systems and guiding enterprises through the entire development lifecycle – from inception and initial testing to deployment and beyond. Where many agencies and freelancers consider their job done at launch, I am just getting started.

I obsess over building systems not just AI components.
➜ optimizing on a system not just individual parts
➜ multiple simple AI modules can bring better performance than one complex model
➜ building feedback loops
➜ building data assets for future leverage
➜ make it easy to swap out components

My approach is data-driven and iterative. I continuously collect and analyze feedback on performance, user engagement, costs produced, and costs saved to refine and optimize AI solutions. This ongoing process ensures that the systems I build deliver real, measurable value long after their initial deployment.