Lead AI/ML Engineer

Lead Generative AI and predictive-maintenance initiatives for gas compressor optimization, delivering solutions from proof of concept through production operation. Architected and shipped a production RAG platform using Azure OpenAI, AWS, LLMs, and Model Context Protocol for context-aware retrieval across technical documentation and industrial sensor data.

Established the LLMOps practice, including prompt and model versioning, MLflow experiment tracking, release evaluation gates, and production-answer review. Built real-time gas-turbine anomaly detection on AWS, designed secure scalable ML APIs, improved LLM request cost through caching and routing, and led international engineering teams through Agile/Scrum and SAFe delivery.

Served as technical lead for a patent IDS analytics platform, designing pattern-recognition and NLP pipelines to clean, normalize, and classify several hundred thousand patent documents.

Replaced manual analyst review steps with text-classification and predictive models, significantly reducing document handling time. Delivered executive analytics dashboards using AWS QuickSight, introduced code-review and unit-testing standards, and mentored junior engineers.

Engineered and deployed machine-learning models on AWS for data-security use cases. Owned large-scale web scraping, text extraction, NLP cleaning, and data-ingestion pipelines supporting model training.

Restructured a large Python codebase into modular, object-oriented components. Removed approximately 30% of duplicated code through improved software design patterns and maintainability practices.

Automated network and operating-system diagnostics using object-oriented Python. Analyzed infrastructure telemetry to identify recurring failure patterns.

Supported the identification of permanent fixes for recurring infrastructure issues through automated diagnostics and data-driven analysis.

Applied Python-based data analytics to diagnose network infrastructure anomalies and investigate technical issues.

Provided escalation-level technical troubleshooting and support for network infrastructure incidents.

First Division (64.57%). Recognised in the Anabin database as a four-year equivalent engineering qualification for EU Blue Card purposes.

Three-year full-time diploma program.