Hello, I'm

EleniPartakki

I work on semantics, data, and enterprise AI — specifically on the part everyone skips: making sure teams actually mean the same thing when they use the same words. It sounds deceptively simple. It almost never is.

Cypriot by birth, Romanian by roots, Londoner by choice — and very fond of a good meal.

scroll

Who I am

Precise,
warm,
and usually right.

🌊 Nicosia, Cyprus 🏰 Transylvanian roots 🇬🇧 London Solutions Engineering Tamr EMEA & APAC CS + Political Science

Most AI systems don't fail because the models are bad. They fail because meaning is inconsistent — different teams using the same words to mean different things, different words to mean the same thing. That semantic drift compounds quietly until automation becomes unreliable and analytics become political.

I work at the intersection of ontology, data engineering, and enterprise AI — designing semantic layers that make business logic explicit, inspectable, and reusable. Ideally without forcing anyone to migrate their databases or learn OWL.

I lead Solutions Engineering for EMEA and APAC at Tamr, which means I spend a lot of time translating between people: ontologists, data engineers, AI engineers, executives, and delivery teams. I'm comfortable in technical depth and equally comfortable in a room where no one agrees on the vocabulary yet.

I genuinely like helping people think through hard problems. I'll stay on a call longer than I should. I'll follow up when I said I would. I care about doing the thing well, not just getting it done.

"I grew up in the last divided capital in the world. I learned early that a shared vocabulary isn't a given — it's something you build, carefully, together."

What I do

Areas of
focus

Six themes I keep coming back to — whether I'm in a workshop, a technical design session, or writing at midnight because an idea wouldn't wait.

🧩

Semantic Infrastructure

Ontology as structural backbone, not academic decoration. Formalising business logic outside of prompts so that AI systems behave consistently — not just most of the time.

🏛️

Enterprise AI Governance

Deployment models, identity context, access control, execution boundaries. Systems that respect governance constraints without just burying ambiguity one layer deeper.

🔗

Solutions Engineering

Leading SE across EMEA and APAC at Tamr — spanning healthcare, life sciences, financial services, manufacturing, and retail. Every sector has its own vocabulary for the same problems.

💬

Narrative & Positioning

How a system is understood, adopted, and trusted is as important as how it works. I spend as much time on the language around a product as the product itself.

⚙️

Data Platforms

Integrating semantic layers with platforms like Databricks without breaking enterprise constraints. Infrastructure-agnostic execution that still holds at scale.

🤝

Human–Machine Alignment

Research background in human-machine interaction and online moderation — now applied to how AI systems earn and maintain trust from the people who actually have to use them daily.

What I'm thinking about

Current
preoccupations

The questions that don't resolve neatly. The ones I'm still working through.

Why do semantic systems succeed when mandated ones fail?

It's rarely because someone had the authority to enforce a standard. It's because someone made the system useful enough that teams stopped fighting it. Adoption is a design problem, not a governance problem.

Can ontology be a deterministic backbone for AI — not just a knowledge graph?

Not something you query when you need a fact, but a structural constraint that makes AI outputs predictable across contexts, teams, and time. That's the architecture I keep returning to.

What does shared meaning actually cost an organisation?

Months of repeated reinterpretation. Political analytics. AI you can't explain to an auditor. The real cost isn't in the tooling — it's in the friction that nobody measured because it looked like "just how things are."

Where does top-down buy-in help, and where does it kill momentum?

There's a tension between the grassroots team that actually uses a semantic layer and the executive who needs to approve it. Getting that sequencing right is an underrated skill in enterprise AI work.

Where I came from

A slightly longer
introduction

I never quite fit into a single category. I've found that useful.

Nicosia, Cyprus

Growing up in the last divided capital in the world

I grew up next to soldiers and UN checkpoints without fully understanding why for the first few years of my life. My father farmed; my mother worked in a grocery store. Money was inconsistent. Stability was earned, not assumed. At six, someone passed down an old computer and that was that — I disappeared into it and didn't really come back.

Community & Politics

Advocating before I had a title for it

Before I had any idea what a career was, I was involved with the European Union and United Nations, representing youth voices on the Cyprus conflict. Somewhere between cycling home from late shifts at a tavern called Palm Tree and studying until 2am, I became someone who genuinely believed that the right conversation, at the right time, changes things.

Smith College, Massachusetts

First in my family to go to university — on a full scholarship

Computer Science and Political Science. The combination made sense to me before I could articulate why: systems thinking, human behaviour, power structures, and the logic underneath everything. My research focused on human-machine interaction and online moderation. I found the friction between what AI systems do and what people believe they do genuinely fascinating — and I still do.

Romania (Transylvania) & Cyprus

Two heritages, both involving a lot of good food and historical complexity

My father's roots are Transylvanian — old forests, castle country, a very specific kind of dry humour. My mother's side is coastal Cypriot — sun, tavernas, warmth that borders on aggressive hospitality. I live somewhere in between. Both taught me to hold complexity without needing to resolve it immediately.

Now — London

Leading Solutions Engineering at Tamr, thinking about semantic infrastructure

I work across EMEA and APAC, which means unusual hours but also genuinely interesting problems at the intersection of enterprise data, AI, and the humans navigating both. I write occasionally on Medium. I track macros with more discipline than is probably healthy. I have strong opinions about eggs.

Outside the work

Also a
person

I take care of the people around me. I take care of my body the same way I approach systems — with structure, but without punishing rigidity. I find joy in small things that are done with real attention.

I'm competitive with myself, not with others. I want things to be good because I made them good — not because I beat someone else at it. That shows up in how I work, how I cook, and occasionally in how I feel about my LDL levels.

🥚

Food, deliberately

I care about how a plate looks and what it means. I have opinions about courgettes that I will share unprompted.

📚

Relentlessly curious

Human brain, geopolitics, economic systems, new technology. I read widely and connect things sideways.

🌿

Outside, often

The outdoors is where I think best. Nature doesn't have an agenda and I find that restful.

🔧

Tinkerer at heart

Gadgets, electronics, things that can be taken apart and understood. Curiosity as a lifestyle.

Get in touch

Let's think
through something together.

I'm particularly interested in conversations about enterprise AI governance, semantic infrastructure, and the gap between what systems can do and what organisations actually manage to adopt. Also happy to talk about Cyprus, Romanian folklore, or the correct way to make eggs.