AI Researcher and Engineer specializing in privacy-preserving machine learning, multimodal AI, and production-scale ML systems. Currently pursuing MS in Computer Engineering at NYU with focus on federated learning, secure computation, and healthcare AI applications.
Core Expertise: MLOps β’ Computer Vision β’ NLP β’ Privacy-Preserving ML β’ Distributed Systems
|
π Privacy-Preserving ML
|
π₯ Medical AI
|
|
π‘οΈ LLM Security
|
β‘ MLOps & Infrastructure
|
π New York, NY | π August 2025 β Present
- π‘οΈ Designing a privacy-preserving distributed data aggregation framework using secure multi-party computation (MPC) and additive secret sharing for federated, privacy-compliant LLM training and evaluation
- ποΈ Developed a two-tier network architecture (heavy/light nodes) with threshold-based secure aggregation, integrating SQL, Oracle RDBMS, and data warehouse systems into privacy-first pipelines
- π Co-authored research paper submitted to ICLR 2026 on secure distributed computation for federated learning
π New York, NY | π Current
- π§ Leading multimodal AI research for medical imaging and automated pathology report generation
- β‘ Built PyTorch pipelines reducing preprocessing time by 20% on NYU HPC systems
- π Demonstrated superior performance of ViLT over MedCLIP in multimodal alignment tasks
π India
- π‘οΈ Integrated anomaly detection algorithms into Google Chronicle SIEM serving 500K+ users
- π Deployed 10+ ML-enhanced log parsers, boosting data usability by 30%
- β±οΈ Reduced manual log management by 30% and alert resolution time by 40%
π¬ IBM Research | Research Intern
- π Developed differential privacy frameworks achieving 35% privacy enhancement
- π Preserved 90% data utility while maintaining strong privacy guarantees
π€ Tech Mahindra | AI Engineer
- π¬ Built NLP chatbots and computer vision solutions
- π Achieved 25% accuracy improvement in production models
Languages & Core
AI/ML Frameworks
MLOps & DevOps
Cloud Platforms
Data & Analytics
Complete MLOps pipeline with distributed training on GPU infrastructure. Provisioned resources on Chameleon Cloud with IaC, implemented multi-GPU training with hyperparameter optimization, and established CI/CD pipeline using GitHub Actions and Argo Workflows.
Modular framework for adversarial testing of GPT models with real-time vulnerability detection. Reduced jailbreak success rates through iterative optimization and automated testing pipelines.
π Amazon Sambhav Hackathon Finalist β’ 5,000+ Active Users
RAG-powered application for export compliance with vector database semantic search and natural language querying of regulatory data.
πΌοΈ AI-Driven Image Categorization
Cloud-deployed image classification system with automated categorization for large-scale processing and production-ready computer vision pipeline.
| Project | Tech Stack | Impact |
|---|---|---|
| π IBM Differential Privacy | IBM Privacy Lib, Python | 35% privacy enhancement with 90% data utility preservation |
| π MSMARCO Search Engine | Information Retrieval | Efficient search algorithms and ranking mechanisms |
| π Cover Letter Generator | LangChain, NLP | Automated personalized content generation |
| βΈοΈ Kubernetes ML Deployment | Kubernetes, MLOps | Scalable ML infrastructure for production |
| π§ Deep Learning Research | Deep Learning | Advanced neural network architectures |
| π» Algorithm Solutions | Python, Algorithms | Optimized competitive programming solutions |
MS Computer Engineering | New York University | 2023 - 2025
Focus: Artificial Intelligence, Machine Learning, Distributed Systems
BTech Computer Engineering | Vishwakarma Institute of Technology | 2018 - 2022
Core CS Fundamentals
Achievements
- π₯ Finalist - Amazon Sambhav Hackathon 2024
- π― Best Attending Team - Hack NYU
- π Best Paper Award - IEEE Pune 2022
Publications & Research
- π IEEE Conference Paper: "Vehicle Characteristics Recognition by Appearance" (2022)
- π¬ ICLR 2026 Submission: "Secure Distributed Computation for Federated Learning" (Co-author)
- π Patent: "CONCRETE GAN: Hybrid AI-based Data Generation Model" (Patent No. 202221001110A)
I'm always interested in collaborating on innovative AI projects, discussing research ideas, or exploring opportunities in privacy-preserving ML and distributed systems.
