CV

MSC. in Information Science (By Research)

Nara Institute of Science and Technology, Japan

Thesis title: Performance Challenges in Containerized Networking: Latency, Jitter, and Bandwidth Beyond 5G.
Laboratory for Cyber Resilience

Bachelor of Technology

Tezpur Central University, India

General 4-year degree in Computer Science and Engineering

Researcher

iPlab, Nara Institute of Science and Technology, Japan

I am conducting research on low latency/jitter user space packet processing within containerized network functions, aiming to advance the next generation of compute interconnect technology.

Research Assistance

ICSCoE, IPA, Ministry of Economy, Trade and Industry, Japan

I conducted research on enhancing core network technology to optimize performance using generic hardware processor architecture for network function virtualization.

Remote Staff Engineer

Ovvy AI: Real Estate Photography, United States

Lead the MLOps team for real estate image editing. We developed fundamental AI models and applied theoretical knowledge and state-of-the-art computer vision advancements to automate image editing. My responsibilities were to conduct research training on custom AI models for superior performance and oversee the deployment of these models to serve millions of requests.

Visiting Researcher

Autonomous Distributed Systems Lab, Umeå University, Sweden

I worked on developing fundamental methods for network anomaly detection using machine learning and data science. Worked on multiple research papers with international collaborations. It was supported by WASP project under Knut and Alice Wallenberg Foundation.

Research Intern

HydroSence Research Lab, Dept. of Civil Engineering, IIT Delhi, India

Used Sentinel-1 SAR satellite data to measure Indian river width on Google Earth Engine Plat-form and designed a web dashboard using LeafletJS, AWS, and Django to visualize crowd-sourced landscape data.

Research Intern

Center for Cognitive Computing Lab, IIIT Allahabad, India

Worked on UI design and did a web-based implementation, which involved the deployment of machine learning models using Flask API, handling physical servers, VPN tunneling over proxy, and back-end and front-end coding.

Development of resilient network function

This study aims to improve the Quality of Experience(QoE) of video flows by preferentially allocating bandwidth to video flows in environments where video communication is essential. Investigating the use of reinforcement learning for developing a custom network transport protocol that can provide the expected QoE.

VSI-DDoS attacks detection on edge cloud

Extensive experiments with testbed and benchmark datasets make the proposed approach effective in terms of detection rate and availability of services, outperforming baseline approaches with 0.9% to 3.2% on average