Jeffrey Dick

Hello! I’m a data scientist passionate about building dependable AI systems. As a PhD-trained researcher, I have 15+ years of experience shipping software used by 100+ research groups globally. With expertise spanning AI systems, data engineering, technical leadership, and research methodology, I know how to make AI tools valuable for your organization.

I love working with data, but my favorite part is building systems that are actually useful. That means paying attention at both ends—from clear understanding of stakeholder requirements to well-designed frontends. Improving user satisfaction is why I’m so excited about GenAI systems: they get better through evaluation sets and feedback loops.

I’m currently open to new opportunities. Reach out via email or LinkedIn to discuss how we can work together.

You can also explore my work on GitHub.

Building dependable AI systems

I don’t just build—I prove. Every AI system I create includes comprehensive test suites and evaluation frameworks to ensure reliability under real-world conditions.

My background maintaining research software means I understand the complete development lifecycle. I’ve solved tough problems cleaning and merging data sources, turning theoretical models into production code, and implementing tests that catch failures before users do. I’m now combining that evaluation-first mindset with hands-on expertise in PyTorch, LangGraph, fine-tuning, and RAG architectures.

This unique blend covers both theory and practice. I can architect complex solutions and implement them end-to-end, always with an eye toward proving they work before deployment.

Delivering value through the entire journey

It’s not just about the code—it’s about bridging technical and business worlds to solve actual problems.

  • For early-stage initiatives, I help you move from concept to production systems with evaluation frameworks that prove value from day one.
  • For enterprise deployments, I integrate complex data sources, design custom LLM workflows, and build feedback mechanisms that combat concept drift as requirements evolve.
  • For customer-facing solutions, I listen to stakeholders and prioritize user experience to ensure your AI systems deliver measurable adoption and impact.

My approach is always the same: understand deeply, build deliberately, and prove rigorously.