Prathamesh Pawar
Boston,MA
Summary
Data Scientist familiar with gathering, cleaning and organizing data for use by technical and non-technical personnel. Well-versed in Language and Generative AI models. Advanced understanding of statistical, algebraic and other analytical techniques. Highly organized, motivated and diligent with significant background in NLP
Overview
4
4
years of professional experience
Work History
Data Scientist (Coop)
Amazon Robotics
01.2024 - Current
- Spearheaded a Cross-organizational project to create ML solution to recognize ant categorize safety incidents on AR floors leveraging LLMs and NLP.
- Developed a Classification pipelines to analyze and tag different issues in robot malfunctions for faster resolution leading to automation of 400 manual annotations.
- Mastered AWS CDK kit for efficient and robust delivery of production level code in data pipelines.
Senior Data Scientist
Retailpulse
03.2022 - 07.2022
- Achieved 92% accuracy in identifying fraudulent delivery images using Custom Object Detection & Image Processing algorithm
- Led a 5-member cross-functional team in building an annotated image repository of grocery shelf pictures, driving Retail Analytics.
Data Scientist
Terra Economics & Analytics Lab
09.2020 - 03.2022
- Designed an in-house model employing Affinity Propagation and Clustering techniques, achieving an Adjusted Rand Index Score of 0.93 for search result quality
- Devised a KNN algorithm for reconstructing locality polygons using geographical coordinates, enabling the generation of addresses solely from property coordinates
- Revitalized the tagger handler engine, significantly boosting the efficiency of the TEAL Check tool which led to a remarkable 70% reduction in erroneous search.
Education
Master of Science - Artificial Intelligence
Khoury College of Computer Sciences, Northeastern University
Boston05.2025
Bachelor of Technology - Electronics And Telecommunication
Government College of Engineering Karad
05.2019
Skills
- Pytorch, Tensorflow, Transformers, Language Models
- AWS: Sagemaker, Lambda, CDK Toolkit, EC2, RedShift
- BERT models, Hugging Face, Regex, OpenCV, PyTesseract, Jupyter
- Python, SQL, PSQL, NoSQL, R, Java
Publications
- Clustering of Spell Variations for Proper Nouns Translated from the languages in the Indian Subcontinent (AIRiAL 2023), Github
- Scientific Summarization: Techniques & Challenges for Summarization and Simplification of Scientific Literature (Under Review), Github
- From Jupyter to Earth: An Example of ML Project Used in Real-World Using TensorRT (The StartUp July 2020), Blog
Projects
- Image classification using CUDA Engine: ML application compiled in C++ with improved latency demonstrating an industry level application of Image Classification algorithm using CUDA Engine and Tensorflow, Github
- Custom Multi-class Object Detection: Object detection model with 15 different classes of custom image database 87% accuracy, Github
Timeline
Data Scientist (Coop)
Amazon Robotics
01.2024 - Current
Senior Data Scientist
Retailpulse
03.2022 - 07.2022
Data Scientist
Terra Economics & Analytics Lab
09.2020 - 03.2022
Master of Science - Artificial Intelligence
Khoury College of Computer Sciences, Northeastern University
Bachelor of Technology - Electronics And Telecommunication
Government College of Engineering Karad
Similar Profiles
Jessica StackhouseJessica Stackhouse
Ambassador at Amazon RoboticsAmbassador at Amazon Robotics
Deitossi MendiolaDeitossi Mendiola
Warehouse Fulfillment Associate at Amazon RoboticsWarehouse Fulfillment Associate at Amazon Robotics
Daniel JonesDaniel Jones
Area Manager at Amazon RoboticsArea Manager at Amazon Robotics