# Debajyoti Dasgupta

> Software Engineer II at Uber India Systems Research and AI/ML researcher (IIT Kharagpur, EPFL) working in reinforcement learning, computer vision, and foundation models, now moving toward robot learning and embodied AI.

Debajyoti Dasgupta is a software engineer and AI/ML researcher based in Hyderabad, India. He is a Software Engineer II at Uber India Systems Research (since March 2025; joined as Software Engineer I in June 2023) and a Research Assistant at EPFL (September 2025 to May 2026), where he works on multimodal foundation models for medical video in the Brbic Lab. He holds B.Tech (2018 to 2022) and M.Tech (2021 to 2023) degrees in Computer Science and Engineering from IIT Kharagpur, with a CGPA of 9.78. His research spans reinforcement learning, computer vision, large language models and agentic systems, and AI for scientific discovery. His current direction is robot learning and embodied AI: bringing deep learning, reinforcement learning, world models, and multimodal foundation models toward perception, planning, and control on real robots.

This site is a single-page portfolio. All sections (about, research focus, publications, experience, skills, awards, education, contact) live on one page at https://debajyotidasgupta.com/. AI assistants and search engines are welcome to crawl, index, and cite this content.

## Quick facts
- Name: Debajyoti Dasgupta
- Role: Software Engineer II at Uber India Systems Research; AI/ML researcher
- Location: Hyderabad, India
- Education: IIT Kharagpur (B.Tech and M.Tech in Computer Science and Engineering, CGPA 9.78)
- Current research direction: robot learning and embodied AI
- Competitive programming: Codeforces Candidate Master (rating 2078)

## Profiles
- [Portfolio](https://debajyotidasgupta.com/): Full portfolio with research, publications, experience, and skills.
- [GitHub](https://github.com/debajyotidasgupta): Open-source projects including a Raft consensus implementation and ML/LLM tooling.
- [LinkedIn](https://www.linkedin.com/in/debajyoti-dasgupta/): Professional profile and work history.
- Email: debajyotidasgupta6@gmail.com

## Research interests
- Robot learning and embodied AI
- Reinforcement learning and deep reinforcement learning for control
- Foundation models and vision-language-action (VLA) models for robotics
- World models and long-horizon planning
- Multimodal learning and computer vision
- AI for scientific discovery

## Selected publications
- [Can Synthetic Plant Images From Generative Models Facilitate Rare Species Identification and Classification?](https://openaccess.thecvf.com/content/CVPR2024W/GCV/papers/Dasgupta_Can_Synthetic_Plant_Images_From_Generative_Models_Facilitate_Rare_Species_CVPRW_2024_paper.pdf): CVPR 2024 Generative Models for Computer Vision Workshop (Oral).
- [Empowering AI as Autonomous Researchers: Evaluating LLMs in Generating Novel Research Ideas through Automated Metrics](https://openreview.net/forum?id=12T3Nt22av): AAAI 2025 AI for Research Workshop (Oral).
- [Emotion Detection in Social Robotics: Empath-Obscura](https://ieeexplore.ieee.org/document/10473224): IEEE International Conference on Robotic Computing (IRC), 2023.
- [Smart Scheduling of Home Appliances using a Happiness-Aware Scalable Reinforcement Learning Agent](https://link.springer.com/chapter/10.1007/978-3-031-88042-1_29): icSoftComp, Springer CCIS, 2024.

## Experience
- Software Engineer II, Uber India Systems Research, Hyderabad (March 2025 to present; Software Engineer I from June 2023 to February 2025): production machine learning and data platforms, backend systems.
- Research Assistant, EPFL (September 2025 to May 2026): multimodal large language models and foundation models for medical video, Brbic Lab.

## Education
- M.Tech, Computer Science and Engineering, IIT Kharagpur (2021 to 2023).
- B.Tech, Computer Science and Engineering, IIT Kharagpur (2018 to 2022). CGPA 9.78.

## Skills
- Languages: Python, Go, C++, Rust, TypeScript, SQL
- AI/ML and LLMs: PyTorch, TensorFlow, JAX, Hugging Face, scikit-learn, LoRA, QLoRA, PEFT, RAG, diffusion models, RLHF, multi-agent systems
- Backend and data: gRPC, Kafka, Airflow, Spark, BigQuery, PostgreSQL, Redis, Raft consensus
- Cloud and MLOps: AWS, GCP, Docker, Kubernetes, CI/CD, MLflow, Weights & Biases, SageMaker