MaksymKhabenko
Engineering autonomous AI systems that replace manual operations at scale. Multi-agent orchestration, LLM tool-use pipelines, and production infrastructure designed to close the loop between business intent and execution without human bottlenecks.
Full-cycle solo delivery: AI Engineering · Backend · Frontend · Business Logic · DevOps.
About
My core focus is AI engineering: multi-agent systems, tool-use pipelines, harnesses, evals and benchmarks. Where API access falls short, fine-tuning and training custom models for the task. Computer vision is a standing interest of its own: VLMs, detection and tracking, real-time inference.
Along the way I build tools for the work itself: custom SDKs, MCP servers, agent infrastructure. Full cycle in one pair of hands: architecture, backend, frontend, deployment.
I work mostly with European companies. Compliance and careful data handling by default.
Over the past year I shipped six production systems: lease abstraction for commercial real estate, autonomous e-commerce support, crypto payments with compliance, cross-border recruiting, immigration law, construction tendering. A new domain every time. The same method.
What counts as a breakthrough at the start of the year is legacy by the end of it. Half of my stack didn't exist a year ago, and half of today's won't survive the next. Holding on to tools here costs more than learning new ones. The one thing I invest in seriously: how fast I rebuild my own process.