Azhar Hajdarevic

At Galeyo, I design and deploy ML risk-scoring models for invoice factoring and build LLM-powered applications with LangChain, LangGraph, and LangSmith. I establish reproducible training pipelines, CI/CD, versioning, and production monitoring to make fintech ML solutions reliable and measurable.

Previously, I developed satellite image segmentation models at Atlantbh and voice biomarker pipelines at Ericsson Nikola Tesla, where my work on schizophrenia screening achieved 86% accuracy and contributed to a SIRS Congress 2026 publication. I also built data pipelines and predictive models for SaaS investment decisions at Clario Media.