Expert in developing and deploying high-accuracy, scalable Machine Learning pipelines for predictive modeling and cost estimation.
I specialize in full-cycle ML solutions, from feature engineering and model tuning to real-time deployment via RESTful APIs. I build robust data pipelines, harmonize complex data sources (Microsoft SQL Server, MongoDB), and ensure model stability and efficiency.
Key Deliverables:
* High-fidelity Predictive Models (XGBoost, LightGBM) for business forecasting and estimation.
* Architecting scalable ML pipelines for efficient data ingestion and model training.
* Real-time inference and seamless integration using FastAPI and Python.
* Data extraction from unstructured inputs using LLMs and classic techniques like TF-IDF and Word2Vec.
Differentiator: Achieved a 98% accuracy rate in AI-based cost estimation, reducing process time by 60% through streamlined pipelines. I deliver production-ready, reliable ML solutions.