Hyderabad, Telangana, India
Results-driven Senior Data Scientist with over 8 years of experience specializing in deep learning, computer vision, and natural language processing (NLP). I excel at leveraging advanced AI architectures including CNN, RNN, Generative AI, and leading LLMs (GPT, Gemini, Transformers, Llama, Mistral, Claude Sonnet) to design and deploy innovative machine learning solutions. What I Deliver • End-to-end ML lifecycle management transforming complex data into production-grade AI systems, from data pipelines to scalable deployment on cloud (AWS SageMaker, GCP, Azure ML) and edge platforms • Proven track record in automating critical processes: legal document analysis, industrial defect detection, real-time pedestrian surveillance, and real-time video analytics • Solutions that consistently drive efficiency and improved decision-making Technical Stack Proficient in core MLOps tools and frameworks such as Python, TensorFlow, PyTorch, MLFlow, CI/CD, Git, and Docker, ensuring robust, reproducible, and high-performance AI solutions. 🚀 Looking to implement AI solutions in your organization? Let's discuss how I can help accelerate your data science initiatives.
- Serving as a key member of the Nestle AI R&D team, focusing on building deep learning CNN and RNN model architectures for pet/animal action classification and recognition, aiming to accurately identify various pet actions in daily life environments. - Engage in all stages of the deep learning project lifecycle, including data labelling and cleaning, to ensure high-quality datasets for model training. - Utilize advanced ML frameworks and pipelines for efficient model development, training execution, and deployment. - Implement MLFlow for tracking experiments, managing datasets, and facilitating model versioning to streamline the development process and enhance collaboration among team members. - Oversee the deployment of models on cloud platforms, ensuring scalable and robust performance in real-world applications. - Continuously explore and integrate the latest advancements in AI and machine learning to improve model accuracy and efficiency
- I had the privilege of working in the Applied AI Team, focusing on critical use cases for TECO Energy Inc & ComEd, which is an Exelon Company. My primary responsibilities included developing and implementing Deep Learning models to identify defects on electrical power poles and household energy meters. - In this role, I architected the setup of Deep Learning models, utilizing Machine Learning frameworks and designing end-to-end pipelines for training execution and model deployment. - I also played a pivotal role in validating the applications' behavior, ensuring real-time compatibility for both Edge and cloud AI on various platforms. - This experience has enriched my proficiency in cutting-edge technologies and has allowed me to make substantial contributions in the field of Applied AI and Energy Solutions. I look forward to leveraging my skills and knowledge to drive innovation in future endeavors.
played a pivotal role in R&D and commercial deployments. My core responsibilities included: - Designing and training CNN and RNN architectures for industrial surface defect inspection, achieving over 98 % detection accuracy in production. - Developing a pedestrian surveillance analytics pipeline: creating a lightweight CNN that ran at 30 FPS on NVIDIA Jetson Xavier edge devices, maintaining 95 % detection accuracy. - Leading the “Thermal Image Object Detection” project: pretrained on visible-spectrum datasets and fine-tuned on limited thermal imagery raising accuracy from 75 % to 89 % in low-light, long-range scenarios. - Establishing end-to-end ML pipelines with Git, and MLflow for experiment tracking, dataset versioning, and model monitoring ensuring reproducible research and seamless production rollouts. - Containerizing inference services with Docker and deploying on Kubernetes clusters, achieving > 99 % uptime for real-time AI services. - Mentoring junior engineers on best practices in PyTorch, transfer learning, and edge-AI implementation; authoring internal white papers on “Sensor Fusion for Edge-AI” and “Active Learning in Thermal Imagery.”