Masters in Computer Science | Georgia Institute of Technology
📅 Aug 2023 – Dec 2024 ★ Atlanta, Georgia- Masters in Computer Science, Major in Machine Learning, GPA: 4.0
- TA in the course Machine learning for trading ML4T CS7646 (AO)
Education
Applied Mathematics | Indian Institute of Technology Roorkee
📅 Jul 2015 – May 2020 ★ Roorkee, India- Integrated Masters in Applied Mathematics, CGPA: 8.23
- Associate Coordinator, Training and Placement Office (IIT Roorkee)
- Coordinator, Ravindra Bhawan Sports Meet
- Member, National Sports Organisation (NSO)
Experience
Graduate Teaching Assistant | Georgia Institute of Technology
📅 Aug 2023 – Dec 2024 ★ Atlanta, Georgia, USA- TA in the course Machine learning for trading ML4T CS7646 (AO)
Google Summer of Code 2024 | Accord Project
📅 May 2024 – Aug 2024 ★ Global Event- Designed and implemented an in-context AI assistant, Co-Pilot, for assisting users in creating Concerto models.
- Integrated the AI assistant into the VSCode web-extension, providing intuitive AI-driven suggestions and guidance.
- Engineered an effective RAG pipeline and backend service to aid users in developing Smart legal contracts.
- Details: https://accordproject.org/news/gsoc-2024-model-creation-ai-assistant-co-pilot/
Member of Technical Staff-2 | Adobe
📅 Jan 2022 – Aug 2023- Designed and developed a content management system to cater linguistic and functional testing activities in Adobe
- Implemented deployment pipelines to ensure services are reliably deployed on K8 clusters
- Designed and implemented a Multimodal Incongruence Detection Model for which I won Innovation Award in recently organized Globalization Innovation-a-thon 2023 in Adobe
Member of Technical Staff | Adobe
📅 Aug 2020 – Dec 2021- Developed a web application used by Adobe Localization Team for workflow orchestration of localization tasks
- Contributed in the development of various APIs and microservices for leveraging products testing in Adobe
- Implemented and integrated an NLP transformer-based solution to solve line break issue in the Thai language
Research Intern | Adobe
📅 May 2019 – Jul 2019- Built a data-driven attribution model for B2B market transactions.
- Explored literature based on various types of LSTMs, survival theory, and deep learning for designing the architecture of the attribution model.
Google Summer of Code 2018 | LibreOffice
📅 05 2018 – 08 2018 ★ Global Event- Selected for Google Summer of Code 2018 in the organization LibreOffice as a student developer.
- The work focused on making the notebook bar UI bugs free and solving the issues related to resizing, theming, and other issues with widgets.
Publications
1. Zero-shot learning based cross-lingual sentiment analysis for Sanskrit text with insufficient labeled data | Journal
📅 2022 * Applied Intelligence (APIN)- Novel methodology for analyzing the sentiments portrayed by Sanskrit text has been proposed
- This research was supported by Ministry of Human Resource Development (MHRD) India
- A dataset of Sanskrit and English translation along with their marked sentiment scores has been constructed
- The paper got accepted in Springer Applied Intelligence (APIN) Journal.
2. Enhancing Conditional Image Generation with Explainable Latent Space Manipulation | arXiv
📅 2024 * ArXiv- Proposed a novel approach integrating diffusion models with latent space manipulation and selective attention mechanisms.
- Achieved superior fidelity preservation and alignment with textual descriptions in the Places365 dataset.
- Achieved significant improvements in image fidelity and textual context alignment, as demonstrated by an improved FID score of 6.89 and a 27.98 in CLIP score.
Projects
Predictive and Visualization Tool for Filmmakers (DVA Project)
📅 08 2023 – 12 2023 ★ Georgia Institute of Technology, USA- Developed a data-driven tool for filmmakers, leveraging machine learning to predict movie success and revenue based on key factors: plot, budget, cast, and release date.
- Implemented a real-time genre prediction model using DistilBERT for dynamic user interactions, applied K-means and Gaussian Mixture Models to cluster users by genre preferences for targeted audience insights.
- Introduced innovations, including a real-time predictive choropleth map and audience preference clustering with t-SNE visualization.
BERT-based Question Answering System
📅 08 2023 – 12 2023 ★ Georgia Institute of Technology, USA- Investigated and evaluated the integration of BERT into Question-Answering (QA) frameworks to enhance performance.
- Conducted experiments, fine-tuned the model on the Stanford Question Answering Dataset (SQuAD), and surpassed baseline models in accuracy and contextual understanding, explored model robustness through a comprehensive evaluation on ambiguous queries.
- Achieved notable results with a fine-tuned model, including start accuracy of 0.6194, end accuracy of 0.66687, and an F1 score of 0.7009, demonstrating its potential for real-world applications.
Machine Translation and Zero-shot Sentiment Analysis in Sanskrit | Dissertation
📅 01 2020 – 05 2020 ★ IIT Roorkee, India- Implemented various seq2seq architectures to achieve translation from Sanskrit to English
- A GAN-based strategy is designed to evaluate the quality of translations
- Used various state-of-the-art sentimental analysis models like BERT to obtain sentiment scores of Sanskrit text
- Achieved an accuracy of 80.67 by using attention-based LSTM
Word spotting in Handwritten Documents | Project
📅 05 2017 – 08 2017 ★ IIT Roorkee, India- This Project was done under the valuable guidance of Prof. Partha Pratim Roy.
- Implemented Phoc (Pyramidal histogram of characters) in a region-based convolution network for segmentation-free word spotting
- Achieved an mAP of 91.6 in query-by-example word spotting
Market Analysis of Bitcoins in India | Project
📅 08 2017 – 11 2017 ★ IIT Roorkee, India- This project was done under the valuable guidance of Prof. J.K Nayak
- Analyzed and interpreted the sampled data using IBM SPSS Statistics and Percentage analysis.
- Concluded multiple hypotheses related to bitcoins in India using ANOVA, Chi-Square, and other methodologies needed for hypothesis testing.
Skills and Awards
Awards
- IIT-JEE 2015 Rank: Achieved a rank of 3519 out of 1.5 million students in the highly competitive IIT-JEE examination.
- Summer Undergraduate Research Award: Received the prestigious Summer Undergraduate Research Award in 2017 for outstanding research contributions.
- Inspire Scholarship Holder: Recipient of the Inspire Scholarship, awarded by the Government of India, in recognition of exceptional talent and dedication in the field of research. Enjoyed this scholarship from 2015 to 2020.
- Globalization Hackathon Runner-up: Secured the impressive position of Runner-up in the Globalization Hackathon organized in Adobe in year 2021, showcasing outstanding problem-solving skills and innovative thinking.
- Globalization Hackathon Innovation Award, Adobe: Honored with the prestigious Innovation Award from Adobe in 2023 for exceptional innovation in the field of technology.
Skills
Programming Languages
Expert:
- C++, Python
Intermediate:
- Java, Javascript
Machine Learning
Frameworks:
- Pytorch, Tensorflow
Libraries:
- Sklearn, Pandas, Numpy, OpenCV, NLTk
Web Development
Frontend:
- ReactJS, TypeScript, JS, HTML, CSS
Backend:
- Flask, SpringBoot, Nodejs
Databases:
- MySQL, MongoDB
DevOps
Tools:
- Git, Docker, Kubernetes, Jenkins
Cloud:
- Azure, AWS