I'm Dimitar. I build the backend systems telecoms and banks run on. After hours, I run Bitcoin infrastructure and write a newsletter about it.

Dimitar Zurnov

Dimitar Zurnov

Software Engineer · JVM & Bitcoin

PLV-2024Available

Four-plus years of JVM-heavy engineering on systems that were already load-bearing when I got there: telecom at A1 Austria, wealth management for RBC, internal platforms that release on a schedule. The side projects are where I get to pick the architecture: a Bitcoin explorer and chart set running on my own node, and the newsletter I write off the back of them.

Currently
Software Engineer, A1 Austria Telekom
since Aug 2024
Based in
Plovdiv, Bulgaria · remote

Selected work

Production engagements and projects I build

Server hardware in a data-center rack: illustrative telecom infrastructure imagery, not a screenshot of A1's internal systems.

A1 Austria Telekom

Telecom · current role

Core platform engineering, release ownership, and internal tooling on a high-impact telecom system.

Feature work in Java 8–22 against Oracle, plus the release side of it: I led a team-wide migration from Bitbucket to GitHub and moved the Jenkins pipelines across without a break in CI/CD. More recently I have been building an internal AI assistant that answers questions across our documentation, ADRs, Jira tickets and codebase.

Java 8–22
range maintained in production
Java 22Spring BootOracleDockerOpenShiftKafka
A multi-screen trading terminal: illustrative fintech imagery, not a screenshot of RBC's internal systems.

Royal Bank of Canada

Wealth management · contract

Microservices for a regulated wealth-management stack.

Turning requirements into small Spring Boot services, making the architecture calls for microservices and container orchestration, and holding the line on JUnit coverage in review. Kafka and Kubernetes, real stakeholders, real audit trails.

JavaSpring BootKafkaKubernetesJUnit

An assistant that has actually read the documentation

One of several things I work on at A1 Austria Telekom.

Abstract illustration of an AI network over a circuit board. Illustrative only, not a screenshot of the actual assistant.

The problem

What a team knows is never in one place. It is spread across documentation, architecture decision records, Jira tickets and the codebase itself, and the answer to a question usually needs two or three of those at once. Engineers know where to dig and lose time digging. Product owners and business analysts often do not know a decision was written down at all.

What I built

An internal AI assistant that answers questions by reading across Confluence, Jira and the code repository, rather than making you go and find the right page first. When it doesn't know something, it fetches an answer and holds it for a human to approve before showing it: a human-in-the-loop check that keeps hallucinations down. Once approved, that answer is written into a vector database, so the next person who asks gets a verified answer instead of a freshly-guessed one.

Who it is for

Both halves of the team, which is the harder part. An engineer wants the ADR and the code path. A product owner or business analyst wants the same decision explained without either. The assistant has to serve a question asked in two different vocabularies and land in the right one.

What it's worth

I can't share the specifics of how it's used or what it touches, but tooling like this is already saving the company a five-figure sum a year.

Alongside this I do feature work in Java 8–22 against Oracle, and I led my team’s migration from Bitbucket to GitHub, moving the Jenkins pipelines across without a break in CI/CD. The platform runs on Docker and OpenShift with Kafka between services.

Recommendation

Dimitar consistently goes above and beyond to meet deadlines and deliver high-quality code. A natural team player — ready with ideas, supportive of colleagues, and actively seeking out new technologies. His quick learning ability made his integration seamless and has been a real contributor to our projects’ success. A pleasure to work alongside.
Vadym Skorupskyy, Senior Manager, Application Development, Royal Bank of Canada

What I do

Backend engineering

Java 8–22 and Spring Boot on production systems. REST APIs, JPA/Hibernate, JPQL, Native and Criteria queries: I reach for whichever the problem actually wants. Services that need to stay up, stay fast, and stay tested.

Platforms & release engineering

CI/CD pipelines, Docker, OpenShift, Kubernetes. I led my team's migration from Bitbucket to GitHub and moved the Jenkins pipelines across without breaking CI/CD. Release coordination, branch strategy, migration work: the kind of work that shouldn't be visible when it's done right.

AI-assisted internal tooling

Assistants that answer questions against a team's own documentation, decision records, tickets and code, and answer them for non-engineers too, not just the people who know which repo to open. I am building one of these at A1 now.

About

The engineer

I'm Dimitar Zurnov, a software engineer based in Plovdiv, Bulgaria, with 4+ years of experience building production JVM systems. My enterprise experience spans telecom infrastructure at A1 Austria Telekom and wealth management platforms at Royal Bank of Canada.

I hold an MSc in Software Technologies (specialization: Software Architecture) and a BSc in Computer Science, both from the University of Plovdiv. Outside enterprise work, I build Bitcoin tooling. explore21.com is a live blockchain explorer I run independently, and Surplus Miner matches ASIC power draw to real-time solar output. I specialise in Java, Spring Boot, Kafka, Kubernetes, LLM integration, and CI/CD pipelines.

Off-hours: powerlifting, volleyball, reading the source.

MSc, Software Technologies

Specialization: Software Architecture

University of Plovdiv "Paisii Hilendarski"

BSc, Computer Science

Foundational CS

University of Plovdiv "Paisii Hilendarski"

Technologies
Languages
  • Java (8–22)
  • JavaScript
  • SQL
  • Bash
  • HTML / CSS
Frameworks
  • Spring / Spring Boot
  • Hibernate, JPA
  • React, React Admin
  • Node.js
  • JUnit, Mockito
Bitcoin / Web3
  • Bitcoin Core
  • BitcoinJ
  • ElectrumX
  • Socket.io
  • MQTT
DevOps
  • Docker
  • OpenShift
  • Kubernetes
  • Jenkins
  • GitHub Actions
Messaging & Data
  • Kafka
  • HermesJMS
  • Oracle DB
  • JPQL / Criteria API
  • Swagger / OpenAPI

Get in touch

Building something worth the effort? I'd like to hear about it.

Quickest path

30-minute intro call · No agenda required

Book a call

Write

[email protected]

Usually back within a day.

Best for

Backend-heavy product work. Release & platform engineering. Anything where the JVM meets Bitcoin. Side-project collaborations welcome.