Brown University

M.Sc. Computer Science

Providence, RI | Expected Graduation: May 2026

Relevant Courses: Machine Learning, Computer Graphics (OpenGL, Ray-Tracing, Scene Lighting), Coordinated Mobile Robotics, Deep Learning, Parallel Computing

Teaching Assistant: Machine Learning

PES University

B.Tech. Computer Science and Engineering

Bangalore, India | August 2020 - June 2024

GPA: 3.93/4 · Specialization: Machine Intelligence and Data Science (MIDS)

Relevant Courses: Machine Intelligence, Image Processing and Computer Vision, Data Science, Data Structures and Applications

Teaching Assistant: Image Processing and Computer Vision

Languages

C, C++, Python, Java, MATLAB, SwiftUI

Frameworks & Platforms

Unix/Linux, PyTorch, TensorFlow, Keras, OpenCV, CUDA, MPI, Isaac Sim, Blender, Simulia

Technologies & Domains

Large Language Models (LLMs), Vision-Language(-Action) Models (VLMs/VLAs), Multimodal Reasoning, Prompt Engineering, 2D→3D Scene Understanding, Image Processing, Computer Vision, Robotics, 3D Mesh Reconstruction & Editing

Relevant Coursework

Advanced Computer Graphics, Deep Learning, GPU Parallelism, Coordinated Robotics

Dassault Systèmes Simulia

Visualization SDE Intern

Johnston, RI, USA | May 2025 – Aug 2025

• Developed and fine-tuned ML models including Graph Neural Networks (GNNs) and Physics-Informed Neural Networks (PINNs) to suggest structural and design optimizations from 3D simulation outputs.
• Built few-shot learning pipelines using Prototypical and Siamese Networks to overcome data scarcity, achieving high generalization with minimal samples.
• Integrated client-server API communication for ML-driven simulation insights and accelerated data extraction by 40% via parallelization.

Brown Intelligent Robot Lab

Researcher

Providence, RI, USA | Nov 2024 – Present

• Conducting research on 3D reconstruction and computer vision for robotics using NeRF, Gaussian Splatting, and Voxel Carving.
• Building advanced robot perception models to emulate human tasks, and incorporating Large Language Models (LLMs) for mesh editing in robotic tool simulation.
• Experimentation is conducted on KUKA robots using ISAAC simulation; research from the SIM-MURI project is under submission to RSS 2025.

Hewlett Packard Enterprise (HPE)

Software Engineering Intern (SAP Developer)

Bangalore, India | Feb 2024 – Aug 2024

• Operated SAP systems to streamline business processes, gained experience with SAP S/4 Hana, SAP Ariba, SAP Fieldglass, and explored the potential of future ERP systems on SAP BTP.
• Resolved 50+ errors in billing processes for order invoices across Israel and Australia, ensuring seamless transactions and efficient operations.
• Led the team that conducted comprehensive 15+ Knowledge Transfer sessions for System Integrators, delving into tools and enhancing collaboration.

SimplyFI SoftTech

Artificial Intelligence Intern

Bangalore, India | Jun 2023 – Aug 2023

• Optimized ML models (VOCA, Geneface++) for audio-to-face transformation in large-scale AI projects, achieving 25% real-time performance gains.
• Prototyped chatbot solutions using GPT4All embeddings and OpenAI APIs to enhance system scalability and user interaction efficiency.

Microsoft PES Innovation Lab

Machine Learning Intern and Mentor

Bangalore, India | Aug 2021 – Oct 2021

• Engineered advanced regression models integrated with GNNs, improving throughput prediction accuracy by 30–35% in dense wireless network environments.
• Led ML workshops for junior students and co-authored an IEEE paper on ML-based WLAN throughput prediction.

Quest Informatics

Software Development Intern

Bangalore, India | Aug 2021 - Oct 2021

• Designed mobile app prototypes using SwiftUI, focusing on seamless user experience.

TPML-WLAN: Throughput Prediction Using Machine Learning

Graph Neural Networks for WLAN Throughput Prediction

PES Microsoft Innovation Lab · Aug 2021 - Oct 2021

Designed and implemented GNN-based models to predict system throughput in high-density environments such as malls and stadiums, improving prediction accuracy by 35%. Built pipelines for data cleaning and rapid model development; results published in IEEE.

Few-Shot Learning for 3D Simulations

Prototypical & Siamese Networks for Data-Scarce Environments

Dassault Systèmes · May 2025 - Aug 2025

Developed few-shot learning pipelines using Prototypical and Siamese Networks to enable high generalization from minimal samples. Applied to 3D simulation stress detection tasks (e.g., McLaren “Critical Zone” detection) with GNN and CNN encoders, improving ML inference speed through multi-core parallelization.

GNN + PINN Structural Optimization Models

Physics-Informed & Graph Neural Networks for Design Suggestions

Dassault Systèmes · May 2025 - Aug 2025

Fine-tuned Graph Neural Networks and Physics-Informed Neural Networks to suggest structural and design optimizations directly from 3D simulation outputs. Integrated models with client-server pipelines for real-time feedback in simulation environments.

Generative AI and Human Cognitive Content

LLM Fine-tuning & Evaluation Framework

Research Publication · ICECTASET 2024

Fine-tuned LLaMA models and developed NLP-based scoring mechanisms to assess LLM-generated content accuracy. Built multi-input RAG + scraping pipelines to evaluate cognitive alignment between human and model-generated outputs.

Kairos: GenAI Task Scheduling App

LLM-powered Scheduling Optimization

Independent Project · 2024

Developed a generative AI-powered scheduling assistant integrating GPT-4, LangChain, and SwiftUI. Achieved 20% reduction in daily scheduling conflicts across 50+ test users. Demonstrated real-world application of prompt engineering and AI workflow orchestration.

VOCA & Geneface++ Optimization

Audio-to-Face Transformation Models

SimplyFI SoftTech · Jun 2023 - Aug 2023

Optimized ML models (VOCA, Geneface++) for real-time audio-to-face transformation in large-scale AI systems, achieving 25% improvement in processing accuracy. Built ML pipelines and integrated GPT4All embeddings for chatbot prototypes.

SIM-MURI

LLM + 3D Mesh Editing for Robotic Tool Selection

Brown Robotics Lab · 2024 - Present

Combined Gaussian Splatting and LLMs for 3D mesh generation and editing to enable robots to autonomously select and adapt tools for specific tasks. Implemented using multi-camera capture, ISAAC simulation, and large language models for automatic annotation and mesh manipulation.

Resilient Distributed Data Streaming System

ML-inspired Streaming Infrastructure

Course Project · 2023

Built a fault-tolerant streaming system in Python simulating a mini-Kafka architecture with replication and high availability across brokers, enabling scalable ingestion for downstream ML workloads.

SIM-MURI

LLM-Enhanced Robotic Tool Selection & Mesh Editing

Brown Robotics Lab · 2024 – Present

Conducting 3D reconstruction research using NeRF, Gaussian Splatting, and Voxel Carving to build high-fidelity mesh models for robotic perception. Developed an LLM-powered mesh editing pipeline enabling KUKA robots to understand and manipulate tools for human-like tasks within ISAAC simulation environments.

Advanced Robot Perception Model

Human Task Emulation through 3D Vision

Brown Robotics Lab · 2024 – Present

Built an advanced perception system for robots to emulate normal human tasks by integrating multi-view reconstruction and real-time mesh editing. Incorporated tool shape transformations (scale, shear, translate) to simulate manipulation tasks in physics-based environments.

Wire Connection Authentication

YOLOv8-based CV System for Electricians

Independent Project · 2023

Developed a YOLOv8 computer vision pipeline to automatically validate wire connections for electricians, reducing manual inspection time by 30% and improving reliability in real-world electrical settings.

Cooperative Payload Transport with Crazyflie Drones

Multi-Agent Control and Communication

Academic Project · 2025 - Present

Implemented cooperative control strategies for two Crazyflie 2.1 drones to collaboratively lift payloads exceeding individual capacity. Designed consensus and leader-follower control modes to maintain altitude difference ≤ 5 cm, minimize payload tilt, and ensure stable transport.

Multi-Camera 3D Reconstruction for Robotics

Scene Understanding & Tool Localization

Brown Robotics Lab · 2024

Employed field multi-camera rigs to generate Gaussian Splatting reconstructions, enabling robots to localize and manipulate tools accurately. Integrated this with mesh editing to automatically annotate tool parts for downstream manipulation tasks.

Samvidha Initiative

STEM Education through Robotics Kits

Technical Head · 2021 – 2022

Led a volunteer team to design and distribute sensor-based robotic kits to rural schools. Conducted workshops to teach fundamental robotics concepts, fostering STEM learning in underrepresented communities.

OpenGL FPS Game and Custom Game Engine

Procedural Rendering & AI Integration

Independent Project · 2023

Built a first-person shooter game and custom raster graphics engine using OpenGL. Implemented procedural terrain generation, skybox rendering, and shader-based lighting. Integrated simple AI agents for dynamic gameplay and optimized frame rendering for smooth performance.

Ray-Tracer with Realistic Reflections and Refractions

Physically Based Rendering in C++

Academic Project · 2023

Developed a ray tracer in Qt (C++) to render 3D scenes with physically accurate effects including soft shadows, reflections, and refractions. Focused on performance optimizations and scene realism through recursive ray calculations and material-based lighting models.

3D Reconstruction using Gaussian Splatting and NeRF

High-Fidelity Scene Modeling for Robotics

Brown Robotics Lab · 2024 – Present

Conducting research on 3D reconstruction using Gaussian Splatting, NeRF, and Voxel Carving to generate scalable, high-quality meshes for robotic perception. Applied multi-camera data capture to create dense, accurate splats that can be rendered interactively in simulation environments.

Interactive 3D Simulation Visualization

Real-Time ML-Driven Structural Insights

Dassault Systèmes · May 2025 – Aug 2025

Engineered client-server pipelines to stream ML-driven 3D simulation insights in real time, enabling interactive structural visualization for McLaren “Critical Zone” detection. Achieved 40% reduction in data extraction time through multi-core parallelization, optimizing GPU pipelines for responsive visual analytics.

ISAAC Simulation + Mesh Editing

LLM-Augmented Tool Manipulation

Brown Robotics Lab · 2024

Integrated 3D mesh editing with NVIDIA ISAAC simulation to allow KUKA robots to interact with dynamically modified tool geometries. Used LLMs to annotate and transform mesh parts (scale, shear, translate) in real-time, bridging simulation with intelligent planning.

Scientific Visualization & GPU Parallelism

High-Performance Rendering

Coursework · 2024

Applied CUDA and MPI parallelization techniques to accelerate 3D simulation pipelines for visualization tasks. Explored GPU memory bandwidth, compute kernels, and pipeline optimizations to enable real-time rendering of complex simulation datasets.