Applied Machine Learning Engineer dedicated to building robust, production-grade AI systems. With deep expertise in optimizing deep learning pipelines and deploying scalable solutions, I bridge the gap between theoretical models and real-world utility.
My approach combines rigorous engineering principles with cutting-edge ML research to deliver systems that don't just work in notebooks—they thrive in production environments.
Object detection, image classification, and visual understanding systems for production environments.
Text analysis, RAG applications, and conversational AI agents powered by large language models.
End-to-end pipeline automation, CI/CD for ML, and production deployment strategies.
Model quantization, inference acceleration, and cost reduction for efficient deployment.
Highly efficient binary sentiment classifier fine-tuned on IMDb using DistilBERT with LoRA for parameter-efficient fine-tuning. Achieves ~85-89% accuracy with 40% smaller memory footprint.
MobileNetV2-based image classification deployed as interactive Streamlit app. Optimized for edge deployment with real-time inference capabilities.
Deep learning model for automated brain tumor detection from MRI scans. Custom CNN architecture categorizing into 4 classes with confusion matrix visualization.
Architected "Melody AI," an autonomous data analyst agent processing 100k+ row datasets. Engineered scalable multi-user backend with FastAPI and Supabase, implementing hybrid state persistence and resilient rate-limiting.
Leading development of Computer Vision pipelines for image analysis and gesture recognition. Optimized inference latency by 40% through model quantization and pruning.
Managed technical workshop sessions for 50+ participants. Managed curriculum execution ensuring alignment between theoretical concepts and practical application.
Whether you're looking to build production ML systems, optimize existing pipelines, or discuss AI strategy—I'd love to connect.