Angel Copilot
An open-source personalized Angel Investment Advisor built in Python.
About
I help teams ship reliable AI through evaluation, agent workflows, and governance that accelerates delivery. My work spans GenAI evaluation, AI agents, ML software quality, multi-agent systems, retrieval augmented generation, and production machine learning systems. I also build open-source tools and frameworks, invest in early-stage AI and technology startups, and share field notes through Beyond the Demo.
Currently
AI Tech Lead at Booking.com & Angel Investor
Available for
Selected work
GenAI maturity framework
An open-source Python framework for assessing GenAI quality and maturity.
ML systems best practices
A framework for analyzing machine-learning best practices through their software quality impact and prioritizing implementation for ML systems.
ML quality maturity framework
A structured framework for assessing and improving ML system quality across Accuracy, Effectiveness, Resilience, and Fairness. This work was accepted into the Applied Data Science Track at KDD 2025.
LLM production evaluation
A practical evaluation playbook covering business-aligned metrics, production-like golden datasets, Judge-LLM development, cost-aware monitoring, dashboards, alerting, prompt search, and synthetic data.
AI agent evaluation
A framework for evaluating AI agents through black-box and glass-box analysis, tool specification quality, tool proficiency, baseline comparison, and pass^k consistency measurement.
Work Background