Kiro x Miro Amsterdam Hackathon winner!
First place at the Kiro x Miro hackathon in Amsterdam with SchoolGuard: a jurisdiction-aware AI tool that helps teachers handle bullying cases.
// Senior AI/ML Engineer · Netherlands
Hi, I'm JeffreyPerson. I work on AI and machine learning in production.
You've reached my little corner of the internet. I'm a senior AI and ML Engineer with a focus on MLOps, natural language processing and LLMs. This site has my blog and pet-projects.
What I work on
04 areas01
ML Platforms & MLOps
Building ML platforms teams deploy on, with SageMaker, Databricks, Airflow, MLflow, and Terraform, so models stay maintainable and monitored.
02
Cloud, across the big three
I've been hands-on with AWS, GCP, and Azure, backed by certifications on each. I've helped AWS shape their ML certification exams too.
03
NLP & LLMs
Natural Language Processing is what I feel most comfortable with. Transformers, classification, and search over messy real-world language, now increasingly with LLMs and agentic workflows.
04
Models in production
Recommendation, ranking, and job-enrichment models serving real traffic at scale, with the plumbing to keep them sharp.
From the blog
All posts →
Hackathon project: DadBot, an AI assistant to slay the dad bod (10000 AIdeas submission)
DadBot is an agentic AI health coach built almost entirely serverless on AWS Bedrock, with personalized ML models, for about three dollars a month. Here's how I built it.
Who reviews the robot? On maintaining AI-generated code
AI can write a working system in an afternoon. Keeping that system maintainable is a different job, and it's increasingly the one that matters.
Our MLOps Story: Production-Grade Machine Learning for Twelve Brands
Things we learned building an MLOps platform with limited means at DPG Media: SageMaker, MLflow, Terraform, and a team of two serving a dozen brands.