Professional Journey
// Building the future with AI & Research
AI Researcher Fellow
iHub, IIIT Hyderabad
- Reproduced and implemented research paper code in the field of drug discovery and computational chemistry
- Optimized model architectures to reduce GPU usage by 40%, improving computational efficiency and cost-effectiveness
- Enhanced model capabilities through architectural modifications and performance improvements
- Conducted research on AI applications in pharmaceutical research and chemical compound analysis
Teaching Assistant - ML4Science
- Conducted hands-on workshops on advanced deep learning for PhD, MS, and UG students from STEM backgrounds
- Taught comprehensive sessions on Diffusion Models (DDPM, DDIM) covering theory, implementation, and practical applications in image synthesis
- Delivered CNN architecture workshops focusing on design patterns, transfer learning, and real-world computer vision applications
- Guided students through research methodologies, experiment design, and model optimization techniques for scientific computing
Open Source Contributor
Scikit-learn & PyTorch
- Actively contribute to Scikit-learn, enhancing machine learning algorithms and improving framework functionality
- Implement and optimize PyTorch modules, enhancing deep learning model performance and functionality
- Collaborate with global developer communities to resolve issues, review pull requests, and maintain documentation
- Develop improvements to ML algorithms including preprocessing methods, model implementations, and utility functions
- Create educational content and examples to help new users understand implementation best practices
Oracle APEX Application Trainee
ProwessIQ Information Systems Pvt Ltd
- Developed an AI-powered image recommendation system using EfficientNet and FAISS, improving image retrieval accuracy
- Engineered a seamless Oracle APEX integration, optimizing database interactions and enhancing system efficiency
- Designed and deployed a scalable AI solution for enterprise applications, reducing computational overhead
- Collaborated cross-functionally with data scientists and software engineers to refine model performance and deployment strategies
- Optimized system performance by implementing data preprocessing pipelines, ensuring high-quality model inputs