Petar Milosavljevic

I am a senior backend engineer with 8 years of experience building scalable backend services, APIs, distributed systems, data-heavy platforms, and automation tools. Most of my recent work has been with TypeScript, Node.js, Elixir, PostgreSQL, MongoDB, Redis, Kafka, RabbitMQ, AWS, GraphQL, and REST APIs.

I started my career at Ericsson, working on backend systems and automation for a telecom billing platform. I owned production tools for invoice generation, billing backdating, prepaid purchases, and monitoring, which taught me to care deeply about correctness, reliability, data consistency, and operational stability.

After that, I worked on CRM and operational software for Forex brokers, building backend services with Node.js, TypeScript, MongoDB, Redis, and RabbitMQ. I developed REST APIs, internal microservices, and integrations with MetaTrader 4 and MetaTrader 5, where data synchronization and production troubleshooting were very important.

My biggest professional experience was at Decidr.ai, where I was one of the early backend engineers in a small team building an AI-powered business operating system for different types of organizations. I was involved in most areas of backend development as the platform grew, including architecture, APIs, data modeling, workflow automation, event-driven systems, analytics pipelines, integrations, payments, and production reliability.

Some of my main achievements there include leading the backend architecture of a no-code workflow automation platform processing 1K+ jobs per minute, designing backend APIs and GraphQL contracts, building analytics pipelines with Node.js and TypeScript, and implementing change capture and event propagation systems with AWS Lambda, SQS, Kafka, and PostgreSQL. I also owned and improved multiple repositories through production debugging, refactoring, testing, documentation, code reviews, and mentoring engineers on system design tradeoffs.

I enjoy solving complex backend problems around distributed systems, data modeling, APIs, event-driven architectures, performance, reliability, and production stability. I like working close to the product, understanding the real business problem, and turning unclear requirements into practical systems that users can depend on. I prefer small, focused teams where I can take ownership, move fast, and have real impact.

I am also interested in AI, both as a product area and as a practical engineering tool. I have worked on LLM-powered workflows, AI agents, RAG pipelines, embeddings, structured outputs, and AI-supported automation. In daily work, I use tools like Cursor, Claude Code, and OpenAI APIs for development, debugging, refactoring, testing, and documentation, while keeping responsibility for final design decisions and code quality.