Laze Gjorgiev

I've built production data and AI systems for MarketRank.ai, NewFire Global Partners, Joyn GmbH, and investment-focused clients. My work spans financial data extraction, compliance monitoring, trading backtesting, customer data platforms, and large-scale cloud migrations.

At MarketRank.ai, I designed a four-stage financial data pipeline that transforms unstructured investment video content into validated, normalized data. I improved extraction precision from approximately 0.55 to 0.91, increased recall from 0.70 to 0.85, and reduced per-video LLM costs by about 47% through prompt caching and observability.

I've led migrations of more than 5TB from on-premise and Azure systems to AWS with zero product downtime, and I owned a client data platform processing 50M+ daily events and 300GB+ of data. By optimizing Kafka and redesigning pipelines, I increased throughput by 500% and cut storage costs by 70%.

I enjoy using data and AI to solve complex business problems, from reliable ETL/ELT and streaming infrastructure to RAG systems, LLM evaluation, and production SaaS products. I work hands-on across Python, SQL, cloud platforms, distributed systems, and data architecture from concept through deployment.