Suvaditya Mukherjee

Education

University of Southern California Master of Science - Computer Science (Artificial Intelligence)	August 2024 - July 2026
NMIMS University Bachelor of Technology - Computer Science (Artificial Intelligence)	August 2020 - May 2024

Work Experience

Magnopus LLC Current: Engineer I, R&D Engineering (Part-time) Prev: Engineer I, R&D (Strike Team) (Full-time)	June 2025 - Present Los Angeles, CA, USA
Multi-agent workforce orchestration: Build ADK-based agentic deliberation systems with over 40 tool-integrated, function calling agents working together backed by a vector database to generate artifacts for several projects based on defined policies.
The Wizard of Oz - Sphere: Helping solve broad-scale AI performance issues related to the project of releasing the original Wizard of Oz movie from 1939 at the Sphere in Las Vegas, as a collaborative project between Magnopus, Google DeepMind, Google Cloud, Warner Bros, and Sphere Studios.
LoRA Training and AI Pipeline setup: Set up end-to-end pipelines to train LoRAs for T2I, I2I, I2V workflows for Wan2.1 VACE using proprietary image and video data.
Video Generation using Diffusion Models: Fine-tuning and optimizing models like Wan2.1 VACE, Hunyuan-DiT, Flux etc., to create new visual generations through tools like PyTorch and ComfyUI.
USC Institute of Creative Technologies Machine Learning Student Worker - Learning Sciences Lab (Part-time)	September 2024 - Present Los Angeles, CA, USA
Course Generation using Generative AI: Leverage Generative AI with LangChain and OpenAI to help make novel techniques for course generation, tutoring content generation, and OpenTutor courses to learn and teach AI for the AIRCOEE program in collaboration with the US Department of Defense, under Prof. (Dr.) Benjamin Nye.
Cogeneration Testbed: Maintain technologies for co-generation of tutoring content using open and cloud-based LLMs to help educators.
USC School of Cinematic Arts Machine Learning Assistant - Interactive Games Division (Part-time)	September 2024 - Present Los Angeles, CA, USA
Student Worker: Assist Prof. (Dr.) Mark Bolas to develop an introductory Python Programming course for Game Developers.
ML Research: Find new approaches to apply Generative AI based on LLMs and Diffusion Models to solve problems at large-scale in Creative Media, with solutions such as generating scripts and summaries based on videos.
HARMAN International Machine Learning Intern (Full-time)	December 2023 - May 2024 Bengaluru, India
K-Shot Rotation-Invariant Object Detection Pipeline Development: Produced new Intellectual Property towards achieving a robust pipeline to perform K-shot object detection without dependence on rotation alignment. Improved pipeline with 35% better results on client data
Zero-shot Time-Series Forecasting with LLMs: Researched on how to achieve zero-shot time-series forecasting through LLMs while building on previous developments.
Spot Instance Handler using Agentic LLMs: Built an agent-based LLM system on Gemini 1.5 Pro and LangChain to help reduce costs by 10% incurred, by running non-critical workloads on spot-instances
Center for Visual Information Technology, IIIT-Hyderabad Research Intern (Full-time)	June 2023 - November 2023 Hyderabad, India
Research: Contributed towards research along Domain Adaptation problems in Autonomous Driving under Prof. C.V. Jawahar and Prof. Shankar Gangisetty
Code Implementations: Operated with internal tools to execute large-scale GPU training and experimentation on Image Segmentation problems
UnifyAI (Ivy) ML Research Engineer Intern (Full-time)	January 2023 - July 2023 London, United Kingdom
Demos and Examples: Developed new demos, examples, and guides to internal and external official documentation, most notably around converting torchvision models into TFLite. Also helped in establishing programs and managing the Google Summer of Code program as an Organization Admin
Internal AI Developer: Prototyped an AI Developer (Code-LLM) to automate and builds upon existing codebases and speeds up internal development, along with handling self-training through Cloud resources such as GCP and AWS

Publications

An overview of the landscape of 3D Generation with PyTorch*

Mukherjee, S. (2025, October 22). An overview of the landscape of 3D generation with PyTorch. https://bit.ly/overview-3d-gen-pytorch-poster

ICPR 2024 Competition on Rider Intention Prediction*

Gangisetty, S. et al. (2025). ICPR 2024 Competition on Rider Intention Prediction. In: Antonacopoulos, A., Chaudhuri, S., Chellappa, R., Liu, CL., Bhattacharya, S., Pal, U. (eds) Pattern Recognition. Competitions. ICPR 2024. Lecture Notes in Computer Science, vol 15334. Springer, Cham.

Sharpen Your Vision: Super-Resolution of CCTV Images Using Hugging Face Diffusers

Mukherjee, S. “Sharpen Your Vision: Super-Resolution of CCTV Images Using Hugging Face Diffusers,” PyImageSearch, P. Chugh, A. R. Gosthipaty, S. Huot, K. Kidriavsteva, and R. Raha, eds., 2024, https://pyimg.co/eoubf

Unlocking Image Clarity: A Comprehensive Guide to Super-Resolution Techniques

Mukherjee, S. “Unlocking Image Clarity: A Comprehensive Guide to Super-Resolution Techniques,” PyImageSearch, P. Chugh, A. R. Gosthipaty, S. Huot, K. Kidriavsteva, and R. Raha, eds., 2024, https://pyimg.co/w4kr8

Project Lingua Franca: Democratizing Information through Unified Optical Character Recognition and Neural Machine Translation*

S. S. Mukherjee, A. Chandra, K. Chemburkar and V. Kulkarni, "Project Lingua Franca: Democratizing Information through Unified Optical Character Recognition and Neural Machine Translation," 2023 International Conference on Modeling, Simulation & Intelligent Computing (MoSICom), Dubai, United Arab Emirates, 2023, pp. 486-491

Guiding the Student’s Learning Curve: Augmenting Knowledge Distillation with Insights from GradCAM*

S. S. Mukherjee, D. Vinod Chandan and S. Dongre, "Guiding the Student’s Learning Curve: Augmenting Knowledge Distillation with Insights from GradCAM," 2023 International Conference on Modeling, Simulation & Intelligent Computing (MoSICom), Dubai, United Arab Emirates, 2023, pp. 587-591

Leadership, Mentorship, and Service

Google Developer Expert for AI, Google Cloud
GDG MAD Mumbai co-organizer (past)
Google Summer of Code 2023 Mentor & Organization Admin (Unify.ai) (past)
Google Developer Student Clubs (now known as GDG on Campus) Lead (past)
Mentor for AthenaHacks USC 2025 - Los Angeles, CA, USA
Judge for DSC Hyphen - Mumbai, India

Awards and Recognition

TensorFlow Community Spotlight Award - Official Keras.io tutorial on Temporal Latent Bottleneck (joint work with Aritra Roy Gosthipaty)
Nominated by TFUG Mumbai to complete the Deep Learning Specialization on Coursera as part of an invitation to the Google Developers ML Bootcamp India.
Runners-up at IET Hack n’ Code 2022 - Mumbai, India
Qualified for INSPIRE scholarship - Govt. of India
Best Student Award - Jankidevi Public School, Mumbai, India