Neelesh Kumar

Senior AI/ML Engineer

Leading end-to-end design of an agentic AI data extraction platform automating Aerodrome Operating Minima calculations from AIP charts, achieving 98.7% extraction accuracy. Developing comprehensive developer documentation and training cross-functional teams to strengthen adoption of AI-driven features.

Agentic AIGenerative AIData ExtractionTechnical DocumentationAI AdoptionTraining

Senior Data Scientist

Engineered and productionized FliteDeck Advisor for optimal Cost Index selection, delivering 1.5–2% fuel savings per flight and $6M annual revenue. Developed a hybrid CNN–Transformer aviation chart extraction engine, saving 36K processing hours annually, and deployed time-series-based altitude and wind optimization models to improve fleet fuel efficiency.

Machine LearningRecommendation SystemsCNNTransformersTime SeriesDatabricksPySparkTableauMentorshipGenAI

Machine Learning Engineer

Honeywell Technology Solutions Lab

Built Hadoop and PySpark pipelines processing ~4TB of flight data per aircraft for large-scale analytics and ML workloads. Delivered 20+ zero-defect production automation solutions, developed a Python-based compression utility reducing build cycle time by 99%, and designed a DNN-based recommendation system for aircraft-to-route assignments.

HadoopPySparkPythonAutomationTableauDeep LearningRecommendation SystemsFlight Safety