From Belgrade to Amsterdam, from writing code to drawing the boxes on the whiteboard.

Started writing code in Belgrade in 2010 — Java, JavaScript, Python, whatever the project needed. Spent seven years moving from individual contributor to tech lead, shipping full-stack solutions across healthcare, education, and e-commerce with teams of up to twenty engineers.

In 2017 I founded Eontek, a technology consultancy that won enterprise clients across EMEA and North America through technical credibility and consultative selling. We maintained 100% client retention before the company was acquired by Sugar Partner. That experience taught me the business side of technology — P&L ownership, go-to-market strategy, and the art of translating engineering capability into commercial value.

From there I moved into solutions architecture — first at HODINKEE in Atlanta (luxury watch marketplace, 6 Scrum teams), then at EPAM Systems in Amsterdam (enterprise transformation, 50+ engineers), and now at MongoDB where I help enterprises modernise their data platforms — from schema design and migration strategy through to AI-powered applications.

At MongoDB I partner with enterprise, FSI, and Digital Native accounts across Benelux and the Nordics to solve their hardest data problems. That means database evaluation and competitive displacement, data modeling and schema design, performance optimisation, migration strategy from legacy systems, and — increasingly — production AI architectures like RAG pipelines and vector search. I run hands-on technical workshops and PoCs across all of these areas, roughly twice a month.

The work is equal parts engineering and storytelling. I partner with CTOs, VPs of Engineering, and Heads of Data to translate platform capabilities into strategic business value — whether that is consolidating five databases onto one, cutting infrastructure costs through Atlas migration, or standing up a production AI pipeline. I build TCO/ROI models, identify and develop technical champions within accounts, and present modernisation roadmaps that customer leadership actually adopts.

I sit at the boundary between architecture and implementation. The data models, migration plans, and production architectures I design for customers are not just slides — I build working prototypes, benchmark them against real workloads, and hand over production-ready reference architectures. Whether it is a schema redesign for a high-throughput transactional system or a RAG pipeline for semantic search, the deliverable is working code. This site is itself a working example: Next.js 16, React 19, MongoDB Atlas, TypeScript end to end.

My technical range spans data platform architecture (MongoDB Atlas, data modeling, aggregation pipelines, change streams, microservices, event-driven systems) through AI/LLM integration (RAG, vector search, LangChain, agentic workflows) to cloud infrastructure (AWS, Azure, Kubernetes, Terraform). I still write code daily — TypeScript, PHP, and Rust.

Every engagement follows a repeatable arc. It starts with discovery — understanding the customer's data landscape, existing architecture, and business objectives. Then I run a hands-on architecture workshop where we sketch the target state together, on whiteboards not in decks.

From there we build a proof of concept — working code, real data, measurable results. The PoC might be a schema migration from a relational database, a performance benchmark against a competitor, or a RAG pipeline for enterprise search — whatever proves the thesis fastest. I wrap it in a value engineering framework (ROI/TCO models, MEDDPICC qualification) that gives the economic buyer the numbers they need to approve production deployment. The cycle typically runs four to eight weeks from first meeting to production decision.