INDEPENDENT TECHNICAL LEADERSHIP
I’m Radu Sora. I work with founders and executive teams to figure out what needs to happen, then take responsibility for making it work.
Healthcare data, knowledge graphs and the infrastructure behind them.
Independent consultancy.
Working directly with you.
01 / SELECTED WORK
Some problems need a few weeks of focused work. Others benefit from years of accumulated context. I’ve done both.
Definitive Healthcare
During a five-year engagement with Definitive Healthcare, I led their Azure-to-AWS migration and carried out most of the implementation myself. It took roughly five weeks and saved the business approximately $2 million in annual infrastructure costs.
Part of a five-year working relationship.
Definitive Healthcare & Forian
My work across these businesses has included US claims processing, patient and provider identity resolution, reference data, affiliations and key opinion leaders. It brings together data engineering, domain knowledge and the practical constraints around sensitive data.
HEALTHCARE DATA / IDENTITY / PLATFORMSProagrica · LexisNexis
I built ontologies for Proagrica, working with RDF and SPARQL. That experience still informs how I approach data products: understand the concepts and relationships in the domain, and make that understanding part of the system.
ONTOLOGIES / RDF / SPARQL02 / WHERE I CAN HELP
I tend to follow a problem across disciplines. These are the areas where I’ve spent a lot of time getting my hands dirty.
I’ve spent years working with US claims data and the problems that come with it: resolving identities, connecting providers and organisations, and turning those relationships into useful products.
What does a record represent? How does it relate to everything else? I’ve built ontologies and worked with knowledge graphs to make those answers explicit. These questions also matter when building useful AI systems.
I care about what a platform costs to run, how it behaves under load, and whether the team can look after it. I work through the architecture, test the assumptions and get involved in the implementation.
03 / HOW I WORK
I work directly with founders and senior executives, with the freedom to find problems, question assumptions and do something about them. I stay close to the technology and lead by example.
Some of my best work has come from long relationships. Understanding the business over time helps me spot things that are easy to miss from the outside.
I start with the business, the data and the people doing the work. Sometimes the most useful contribution is noticing a problem before it becomes an expensive commitment.
I want to understand why a design makes sense. That means looking at evidence, running benchmarks and counting the cost of running it. A fashionable tool still has to earn its place.
I design, write code, work through the awkward details and help the team move forward. I explain the reasoning as we go, so the people looking after the system understand the choices behind it.
04 / A BIT ABOUT ME
I tend to follow my curiosity.
It has proved useful.
I’m based in Cambridge, UK, and have spent over 20 years working with technology businesses across the UK, Silicon Valley and the Nordics. Polysemantic is my independent consultancy. When we work together, you work directly with me.
My background spans healthcare data, media analytics, agriculture, low-latency trading and financial services. My PhD in Engineering at the University of Warwick used semantic technologies to infer viewers’ interests from the online video they watched.
I like getting into subjects that seem interesting, sometimes in more detail than is strictly necessary. Often something from one field turns out to be useful in another. I value independent thinking, direct feedback and people who take responsibility for their work.
05 / WRITING & TALKS
A talk about running Spark on Kubernetes, the economics of managed platforms, and understanding the infrastructure you depend on.
GET IN TOUCH
A difficult technical problem, a data product that could work better, or an idea you think we could build together. I’m happy to have a conversation.