Engineer. Hacker.Vibe Coder. Builder.

Walk on fire, breathe ammonia (2009 – 2019)

Branimir started as an automation engineer at Solvay Sodi, one of Bulgaria's largest chemical plants. Not the kind of job where you sit in front of a screen — the kind where you bail water by hand to keep PLCs alive during a flood, climb 180-metre chimneys in wind, and witness the destructive power of a multi-tonne steam head punch through a metre-thick concrete wall.

He automated compressors, utilities, and process units on both Honeywell and Siemens hardware, ported Honeywell control blocks to Siemens SCL, and cleaned up after consultants who left systems broken. He learned what industrial software actually costs when it fails. When things went wrong, external companies would say: "If Branko is here, everything will be all right." In technical terms: Branimir had achieved bhai status — the one certification no university awards.

Break it until you make it (2019 – 2025)

At Hilscher, Branimir moved into software and discovered Python. He hit a gap no existing library could fill: protocol stacks built ONLY for communication, not for testing. You cannot craft malformed packets. You cannot simulate edge cases the spec does not cover.

That gap is where the work began. An SDK grew to cover DeviceNet, CAN, Modbus, OPC-UA, and MQTT. Brutal deadlines from tier-1 American semiconductor equipment manufacturers. Impossible milestones. All delivered. The model was lean: developer and tester working in parallel, every artifact shipped already tested.

Branimir had a penchant for crashing things on a Friday afternoon, or right before a public holiday — because that is when developers reveal their true dedication. He trained six engineers and earned the nickname the bastard from heaven in tester mode and master-blaster in developer mode.

The best collaboration is through conversation (2025 – Present)

At Heidelberg Materials, Branimir worked on PxTrend — an industrial historian built by Paranext. When the setup stalled, he drove to the nearest plant in Devnya, talked to the operators, learned from the clients, learned from the creator, and built understanding the only way that works — by experimenting, using it, and solving problems. The understanding of what a historian should do became the foundation for what came next.

A special thanks to Edgar Poth, founder of Paranext, for sharing his knowledge directly — shadowing him from 4am to 11pm through installations, upgrades, and live migrations of PxTrend was worth more than any technical book. It was an eye opening experience with a showcase of real problems on the plant floor.