Backend engineer for Java, Spring Boot, Python, Go, and AI. I fix, build, and harden production systems: APIs, data pipelines, RAG, and automation.

I lead a development team at NITES building production systems for healthcare and the energy sector, and on the side I build deep, well-tested systems across the stack: distributed transactions, event pipelines, document AI, automation, and cloud-native infrastructure. I am not a CRUD developer. The interesting problems are the ones where two writes can disagree, events arrive late, or a document is a scan, not text.

What I actually build, with public evidence on github.com/Buljic:

- Spring Boot and Java. Transactional outbox services that commit business state and events atomically to PostgreSQL and publish to Kafka with retries and idempotency. JTA two-phase commit across databases. REST APIs with real tests. Spring Certified Professional 2024.
- Python. FastAPI services, n8n workflows, forecasting and anomaly detection, RAG over real documents with a hand-written hybrid retrieval layer.
- Go. A Kubernetes operator with CRDs, retries, and S3 artifact collection. An async notification service over NATS JetStream delivering through Telegram and signed webhooks.
- TypeScript and React. A tutoring platform with roles, WebSocket chat, and WebRTC signaling. A multi-tenant SaaS API in Node.js and TypeScript with a BullMQ worker.
- Databases. PostgreSQL as the main store, MySQL, Elasticsearch kept in sync through outboxes, Redis caching.

Day job: team lead at NITES on digitalized grid-connection workflows, where business state lives in workflow-engine state machines and integrates with surrounding services. I also hire and mentor engineers there.

Background: BEng in computer science, MSc in computer engineering in progress, published research, Spring Certified Professional 2024, backend engineering and DevOps certifications. Fluent English, direct communication. I use AI coding tools heavily but review, test, and own every change. I join existing codebases without drama and ship focused, verifiable fixes.

Work Terms

Availability: part-time, 10 to 20 hours a week, Europe/Sarajevo time (UTC+1), with some flexibility for your timezone. Fixed-price quotes for small scopes, funded milestones for larger ones.

Communication: clear and direct, updates every working day while a job is active, questions early rather than late, all through Guru messages.

Process: I join the existing codebase directly, read the code before touching it, and ship small verifiable changes with tests where the project has them. I use AI coding tools heavily but review, test, and own every change.