Sourabh Vasant Gothe
Bangalore
Summary
Passionate about working in Machine Learning and Deep Learning, Generative AI. Experienced in C++, Java, Python, Algorithms, Machine Learning, and Deep Learning with a solid theoretical understanding. I have completed my M. Tech in Data Science from the Indian Institute of Science (IISc) with a CGPA of 9.3. I have authored 10+ research papers published in conferences such as WACV, ICASSP, CVIP, and ICSC. Additionally, I actively participate as a reviewer for conferences like WACV, EMNLP and ICASSP. I hold a portfolio of 10+ patents, including 8 A1 grade patents. I am eager to bring my passion for innovation and research to the team.
Overview
8
8
years of professional experience
Work History
Chief Engineer, Machine Learning
SAMSUNG R&D
03.2024 - Current
- Currently leading a research & dev project focused on Knowledge Graphs and Large Language Models related use-cases on mobile device to enhance user personalization experience
- (July- Dec, 2024) Developed AI product for aesthetic widget placement to enable enhanced on-device user experience. I lead a team and built an end-to-end solution, including a vision model (one-stage anchor-free object detection) to identify the appropriate location to place a widget automatically.
Lead Engineer, Machine Learning
SAMSUNG R&D
03.2021 - 01.2024
- Worked on the project of Generic Temporal Segmentation problem, that caters many use-cases
- Developed a highly optimized deep learning architecture with only 3M trainable parameters — 22× fewer than SOTA — achieving a 13.92% accuracy improvement (ICASSP 2023).
- Proposed a novel hybrid algorithm to generalize across diverse scenarios, published in WACV 2024.
- Owned the end-to-end product lifecycle including DNN modeling, unsupervised learning, data curation, and on-device inference optimization.
- Successfully commercialized the solution in production for Samsung Galaxy S24, S25, and other flagship devices.
Senior Software Engineer
SAMSUNG R&D
02.2019 - 03.2021
- Contributed to the Text Intelligence team to enhance mobile user experience through innovative NLP features.
- Developed and deployed a multilingual texting feature tailored for the Indian market, including language detection, code-switching, and macaronic language modeling (top 5 Indian languages)
- Built a complete on-device grammatical error correction system, optimized for mobile devices (ICSC 2021).
- Developed Language Detection (Patent granted, ICSC 2020)and a Lightweight Language Model (ICSC 2019)
- Designed an Emoji Drawing Detection system (Patent granted) for edge devices like wearables; developed an end-to-end native (C++) neural inference engine for CNN-based models (CVIP 2022)
- Successfully commercialized native inference engine and NLP features in Samsung Neural Keyboard across Galaxy flagship devices
Software Engineer
SAMSUNG R&D
07.2017 - 03.2019
- Evaluated multiple language model engines (e.g., PhillIT, XT9, Gboard, SwiftKey) for next-word prediction and auto-correction to assess quality and UX impact.
- Developed an Android-based evaluation framework to automatically benchmark models on large-scale text corpora.
- Gained hands-on experience with advanced data structures like MARISA and CEDAR TRIE for efficient user language modeling
Research Intern
SAMSUNG R&D
01.2017 - 07.2017
- I developed a machine learning algorithm based on HMM that detects human activity using low-resource devices' sensors
Education
M.TECH - Data Science And Business Analytics
Indian Institute Of Science
07.2023
BACHELORS IN ENGINEERING - Computer and Information Sciences
PES Institute of Technology
06.2017
Skills
- Machine Learning
- Deep Learning
- Computer Vision (Image/Video Understanding)
- Natural Language Processing (LM)
- PyTorch
- C
- Python
- Java
- Android/Kotlin
- Innovative & Self-motivated
- Team Work & Leadership
- Communication
Patents
- A1 Grade | System and Method for Intelligent Temporal Segments based on Class Agnostic Video Networks
- A1 Grade | Method And Electronic Device For Predicting Plurality Of Multi-Modal Drawings
- A1 Grade | System and Method for Input Enrichment based on OnDevice Text Boundary Detection
- A1 Grade | System and Method for OnDevice Transcription To provide Multi-modal Input Experience
- A1 Grade | System and Method for Localizing Inconsistent Motion in Video
- A1 Grade | System and Method for Video Generation based on Spatio-temporal concept relationship modelling
- A1 Grade | System and Method for Generating contextual transition based on guided denoising of flow
- A1 Grade | System and Method for content-aware element interactions based on motion estimation
- B1 Grade | Word Variant Disambiguation for Romanized Abugida Scripts
- B1 Grade | Ambiguous Input Algorithm for Texting in Native Abugida Scripts
- B1 Grade | System and Method for Language Detection based Multilingual Experience
Awards
- Samsung Best Paper Awards 2024 | Bronze Award for our paper PhysID
- Samsung Annual Award 2023 | Eminent Author of the Year
- Samsung Excellence Award 2021, 2022, 2023 | Star of the Quarter
- Samsung Software Competency Level, Professional, Achieved this level with a thorough understanding of advanced data structures & algorithms.
Research Publications
- Self-Similarity Is All You Need For Fast And Light-Weight Generic Event Boundary Detection, Sourabh Vasant Gothe, Jayesh Vachhani, Rishabh Khurana, Pranay Kashyap, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2023)
- What's in the Flow? Exploiting Temporal Motion Cues for Unsupervised Generic Event Boundary Detection, Sourabh Vasant Gothe, Vibhav Agarwal, Sourav Ghosh, Jayesh Vachhani, Pranay Kashyap, Accepted in IEEE/CVF Winter Conference on Applications of Computer Vision (WACV 2024)
- SAMNet: Semantic Aware Multimodal Network for Emoji Drawing Classification, Sourabh Vasant Gothe, Rishabh Khurana, Jayesh Vachhani, Rakshith S, Pranay Kashyap, Seventh International Conference on Computer Vision & Image Processing (CVIP 2022)
- Language Detection Engine for Multilingual Texting on Mobile Devices, Sourabh Vasant Gothe, Sourav Ghosh, Sharmila Mani, Bhanodai Guggilla, Ankur Agarwal, Chandramouli Sanchi, 2020 IEEE 14th International Conference on Semantic Computing
- An Efficient System For Grammatical Error Correction On Mobile Devices, Sourabh Vasant Gothe, Sushant Dogra, M. Chandra, Chandramouli Sanchi, Barath Raj Kandur Raja, 2021 IEEE 15th International Conference on Semantic Computing
- Real-Time Optimized N-Gram for Mobile Devices, Sharmila Mani, Sourabh Vasant Gothe, Sourav Ghosh, A. Mishra, Prakhar Kulshreshtha, M. Bhargavi, M. Kumaran, 2019 IEEE 13th International Conference on Semantic Computing
- EdATLAS: An Efficient Disambiguation Algorithm for Texting in Languages with Abugida Scripts, Sourav Ghosh, Sourabh Vasant Gothe, Chandramouli Sanchi, Barath Raj Kandur Raja, 2021 IEEE 15th International Conference on Semantic Computing
- Self Similarity Matrix based CNN Filter Pruning, Rakshith S, Jayesh Vachhani, Sourabh Vasant Gothe, Rishabh Khurana, Seventh International Conference on Computer Vision & Image Processing (CVIP 2022)
- Class Agnostic, On-Device and Privacy Preserving Repetition Counting of Actions from Videos using Similarity Bottleneck, Rishabh Khurana, Jayesh Vachhani, Sourabh Vasant Gothe, Rakshith S, Seventh International Conference on Computer Vision & Image Processing (CVIP 2022)
- SAM-GEBD: Zero-Cost Approach for Generic Event Boundary Detection, Pranay Kashyap, Sourabh Vasant Gothe, Jayesh Vachhani, Vibhav Agarwal, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024)
- PhysID: Physics-Based Interactive Dynamics From A Single-View Image, Sourabh Vasant Gothe, Bronze Award in Samsung Best Paper Award 2024, IEEE International Conference on Acoustics, Speech and Signal Processing ICASSP 2025
Timeline
Chief Engineer, Machine Learning
SAMSUNG R&D
03.2024 - Current
Lead Engineer, Machine Learning
SAMSUNG R&D
03.2021 - 01.2024
Senior Software Engineer
SAMSUNG R&D
02.2019 - 03.2021
Software Engineer
SAMSUNG R&D
07.2017 - 03.2019
Research Intern
SAMSUNG R&D
01.2017 - 07.2017
M.TECH - Data Science And Business Analytics
Indian Institute Of Science
BACHELORS IN ENGINEERING - Computer and Information Sciences
PES Institute of Technology
Similar Profiles
KUMAR HARSHWARDHANKUMAR HARSHWARDHAN
AWS DevOps Engineer at Samsung R&DAWS DevOps Engineer at Samsung R&D
Tejaswini NTejaswini N
Associate - HRBP at Samsung R&DAssociate - HRBP at Samsung R&D
Yash GargYash Garg
Chief Engineer at Samsung R&DChief Engineer at Samsung R&D