Open to Opportunities

Hi, I'm Rithwick Sethi

AI Engineer and researcher at Carnegie Mellon University, building intelligent systems at the intersection of deep learning, generative AI, and autonomous robotics.

Get in Touch View Projects

Research Papers

3.7

QPA at CMU

1st

Dept Rank (DTU)

National Hackathon Winner

About Me

Building the future with AI

I'm a Master's student in AI Engineering at Carnegie Mellon University with a passion for pushing the boundaries of machine learning, deep learning, and generative AI. My journey spans from founding one of India's leading university AI societies to engineering production systems at Apple.

With 7 research publications, a Vice-Chancellor's Gold Medal, and experience across autonomous robotics, 5G systems, and generative video models, I bring a unique blend of academic rigor and industry pragmatism to every project.

Currently, I serve as Head Teaching Assistant for Generative AI & LLMs at CMU, mentoring the next generation of AI engineers while pursuing cutting-edge research.

🏆
Vice-Chancellor's Gold Medalist
Department Rank 1 with 9.59/10 GPA at Delhi Technological University

🍎
Apple SDE
Built ULTRON framework reducing test case generation time by 95%

🚀
$11M Drone Funding
Demonstrated autonomous drone capabilities to the Indian Army

🎓
Head TA at CMU
GenAI & LLMs, Wireless Communications (PhD Level)

Education

Academic journey

Carnegie Mellon University

M.S. in Artificial Intelligence Engineering — ECE

Pittsburgh, PA Dec 2026 QPA 3.7/4.0

Head TA — Generative AI & LLMs, Wireless Communications (PhD Level)
Coursework: Machine Learning, Deep Learning, Gen AI, LLM Systems, Estimation Detection Learning, Systems & Toolchains for AI

Delhi Technological University

B.Tech in Electronics & Communications Engineering, Minor in AI/ML

New Delhi, India 2020 — 2024 GPA 9.59/10 · Rank 1

Vice-Chancellor's Gold Medalist, Branch Gold Medalist, Merit Scholarship Recipient
Two-Time National Winner — Smart India Hackathon (2022, 2023)
Secured $11M funding after demonstrating autonomous drone capabilities to the Indian Army

Skills

Technical toolkit

Languages

Python C MATLAB Verilog Bash SQL

ML / AI Frameworks

PyTorch TensorFlow Keras JAX TinyML OpenCV

LLM / GenAI

RAG LangChain LangGraph MCP LoRA / PEFT

Data & Infrastructure

Spark Kafka Neo4j Docker AWS Git Pandas NumPy

Robotics & Systems

ROS Gazebo PX4 / SITL 5G / LTE 3GPP

Certifications

Deep Learning GANs RAGs Aerial Robotics IoT

Experience

Where I've worked

Apple

Software Development Engineer — Wireless Technologies & Ecosystems

Jul 2024 — Jul 2025

Developed and optimized NAS (Non Access Stratum) L3 pipelines for large-scale validation of cellular modules ensuring reliability across complex 5G/LTE systems per 3GPP Standards.
Debugged and resolved issues using fuzzing, static analysis (Coverity), and large-scale data-driven testing, improving error detection efficiency.
Investigated end-to-end system interactions, from AP triggers to baseband processing, ensuring robust integration across layers.

Apple

Intern — Wireless Technologies & Ecosystems

Jan 2024 — Jul 2024

Built ULTRON, a standardized and automated framework to streamline the tedious and error-prone process of host test case development.
Reduced time in test case generation by 95% (5-6 hours → 10-15 minutes), improving efficiency and reliability.
Integrated ULTRON in Apple's Internal Tools by providing extensible API support for org-wide adoption.

MITACS — École de Technologie Supérieure

Research Intern — AI-based UAV Swarm Coordination

Jun 2023 — Sep 2023

Designed and tested Deep Reinforcement Learning and Heuristic-Based AI models in MATLAB and ROS for optimal path allocation of UAV swarms in low-visibility and uncertain target motion scenarios.
Presented a novel framework in an accepted research paper at ICECCT 2024.

Biometric Research Laboratory — DTU

Research Intern — Human Action Recognition

Jun 2022 — Sep 2022

Created a robust framework for video sampling and action classification using spatial-temporal feature extraction via a STSTB model and EfficientNet-based autoencoder-classifier, tested on the HMDB51 dataset.

Projects

Things I've built

Comic2Video — From Panels to Motion

Generative AI · Carnegie Mellon University · Nov 2025

Fine-tuned WAN-2.1-FLF2V-14B latent video diffusion model using LoRA (PEFT) on A100 GPUs for context-consistent motion interpolation.
Built a 35GB dataset hosted on HuggingFace with multimodal prompt metadata via MoonDream2 LLM.
Integrated RAFT optical-flow auxiliary loss and VRAM-aware loss staging; leveraged AWS, Docker, OpenAI, WandB.
Achieved FVD = 898.36, SSIM = 0.983, LPIPS = 0.0088 with LoRA-32.

PyTorchLoRAHuggingFaceAWSDockerWandB

MOMVO-Based UAV Swarm Trajectory

ICECCT 2024 (Accepted) · May 2024

Introduced a novel MOMVO-based framework for UAV swarm trajectory planning with advanced computer vision for dynamic formation and real-time object recognition.
Validated in SITL/QGC/PX4 simulations, reducing average path length to 120.589 units vs. 128–135 from existing algorithms.

MATLABROSPX4Computer VisionSITL

PCOS Detection via Machine Learning

ICCCNT 2023 (Published) · Jun 2023

Applied and compared ensemble models, dimensionality reduction, and PCA on Kaggle datasets.
Achieved 91.11% accuracy based on visible features alone, enabling early PCOS detection.

PythonScikit-learnPCAEnsemble Methods

StrikeTag: Kamikaze Drone GeoTagging

DAUS Conference 2025 (Accepted) · Dec 2022

Designed and implemented a "Kamikaze Drone" with YOLO-based object detection, GeoTagging for target coordinates, and autonomous strike execution.
Simulated in Gazebo via ROS and SITL.

YOLOROSGazeboSITLPython

Publications

Research contributions

StrikeTag: Object Detection Based GeoTagging For Kamikaze Drone Surveillance

R. Sethi, B. Goel, K. Agrawal, P. Sharma, C. Dhiman, D.K. Vishwakarma

DAUS 2025 First Author Accepted

MOMVO-Based Trajectory Generation for Multi-Target Search By A Swarm of UAVs

R. Sethi, D. Chauhan, T. Nassa, R. Rohilla

ICECCT 2024 First Author Accepted

A Comparative Study on Different Machine Learning Algorithms to Detect PCOS

R. Sethi, D.K. Vishwakarma, R. Ray, S. Ganguly

ICCCNT 2023 First Author Published

AutoBot: A Semi-Autonomous Bot

R. Sethi, D. Chauhan

ICICAT 2023 First Author Published

Audio Based Machine Fault Diagnosis Using Hybrid Feature Extraction and Ensemble Learning

S. Singhal, B. Goel, K. Agrawal, R. Sethi, S. Sah, R. Jain, D.K. Vishwakarma

ICCCNT 2024 Published

KrishiGrow: An Expert System Based Macro and Micronutrient Visualization System for Smart Crop Management

S. Mondal, D. Roy, S. Ganguly, R. Bhowmick, R. Sethi, S. Banerjee, D. Sutar

JAEM 2021 Published

AgroNet: A Decentralized Platform for Collaborative Community-Driven Consultancy for Farmers & Agro-Vendors

S. Mondal, Z. Agar, R. Ray, S. Ganguly, R. Bhowmick, R. Sethi

JAEM 2021 Published

Leadership

Beyond the code

Co-founder & President

AIMS-DTU (Artificial Intelligence & Machine Learning Society) · Sep 2021 — Feb 2024

Founded AIMS-DTU to cultivate an inclusive AI/ML learning space, offering knowledge and support to everyone eager to explore the field.
Grew the society to 1,600+ students in 2 years, building one of the university's leading AI/ML communities through mentorship, webinars, competitions, and published research.