Applied Scientist II · Microsoft
Knowledge graphs & GNN-based threat detection; scalable Azure Synapse ML pipelines for real-time risk scoring.
Applied Scientist II · Microsoft · Bangalore
3+ Years Industry Experience
3+ Years Research Experience
2 Publications (CORE A)
Building graph-based threat detection systems using GNNs and knowledge graphs at Microsoft. Open-source contributor to NetworkX. Prior work includes multi-agent GenAI systems, LLM optimization, and scalable ML pipelines across Myntra and Tekion.
- RLHF
- AI Safety
- Multi-Agent Systems
- Post-Training
- Interpretability
Data Scientist · Myntra
Multi-agent GenAI chatbot (58.86% MAU uplift); CLIP-based recommendations; PySpark pipelines on 500M+ records.
Data Scientist · Tekion
LLM fine-tuning with 98% latency reduction; RAG chatbot for 800+ car models using Langchain.
Research · IIIT Hyderabad
Minimax GAN optimizer (WCCI 2024); novel learning rate method (IJCNN 2023, CORE A); GPU-accelerated PyTorch solvers.
WCCI 2024 · IEEE
A Gauss-Newton Approach for Min-Max Optimization in Generative Adversarial Networks
A novel first-order method for training GANs using a modified Gauss-Newton method to approximate the min-max Hessian. Achieves highest inception score for CIFAR10 among all compared methods.
IJCNN 2023 · IEEE
Angle Based Dynamic Learning Rate for Gradient Descent
A novel adaptive learning rate method using the angle between current and orthogonal gradients. Outperforms state-of-the-art optimizers on CIFAR10/100 across ResNet, DenseNet, EfficientNet, and VGG.
Get in Touch
Open to research collaborations, applied ML roles, and open-source contributions. Feel free to reach out.