Machine Learning Information Technology

Pruthvik Elemati

317-***-**** ********@******.***

Education

Purdue University Indianapolis, USA

Master of Science in Computer Information Science GPA : 3.67/4.0 Jan 2023 – Dec 2024(Expected)

• Relevant Courses : Natural Language Processing, Deep Learning, Data Mining, Statistical Machine Learning Indian Institute of Information Technology Sricity, India Bachelor of Technology in Electronics and Communication Engineering GPA:7.29/10.0 Aug 2017 – June 2021

• Relevant Courses : Statistical Data Analysis, Pattern Recognition, Advanced Statistical Methods, Digital Image Processing

Experience

Graduate Research Assistant NLP, Python, Pytorch, HuggingFace Feb 2023 – Present Purdue University Indianapolis, USA

• Collaborated with Dr. Hyeju Jang on extensive literature reviews, data processing, and analysis tasks in commonsense reasoning and metaphor detection.

• Conducted experiments on injecting commonsense into BERT using ConceptNet and ATOMIC knowledge graphs through adapters and other mechanisms and evaluated their performance on metaphor detection task.

• Evaluated current metaphor detection SOTA models across diverse metaphor types ( conventional, novel, high metaphoricity, low metaphoricity ), providing crucial insights into model performance and guiding future research.

• Fine-tuned multiple language models, and enhanced existing models to assess and improve their performance in metaphor detection and interpretation.

• Working on interpreting metaphors by generating the literal counterparts using knowledge graphs like ConceptNet, ATOMIC.

Senior Engineer-Industrial IoT Python, Angular Oct 2021 – Dec 2022 Larsen and Toubro Infotech Hyderabad, India

• Worked with clients in oil and gas sector, providing them data driven and ML solutions that meet their business objectives.

• Built a Remaining Useful Life (RUL) estimator model and health-index using ML for preventive maintenance of the pump, reducing maintenance costs by 30 percent.

• Developed a machine health monitoring system for analytics and raising alarms in case of rule violation.

• Designed a data cleaning pipeline, detecting anomalies, flatlines, missing values and drift in real-time streaming data.

• Designed Dashboards in Angular with TypeScript to present the KPI’s, notifications and important events.

• Identified Important trends in the data and the influence of specific events on the productivity of overall machinery using time series forecasting and analysis.

• Performed predictive modeling and statistical analysis on the sensor data. Software Engineering Intern React, Node, Python Jan 2021 – May 2021 Edvenswa Tech Hyderabad, India

• Abstracted different parts of quiz application logic into react components.

• Loaded data from back-end to front-end using Node and Express.

• Worked on a tool to extract scores from golf scorecard using image processing and OCR. Projects

Quora Question Pairs Similarity Python, NLP, Pytorch Course : Natural Language Processing Jan 2023 - May 2023

Worked on detecting if a pair of questions in the dataset are duplicates.

The project was undertaken with an aim to reduce redudancy and declutter the platform.

Came up with a Multi-task learning model to reduce the overfitting problem of existing Sentence BERT.

Tested the generalizability of the model on Stack Overflow Questions. 3D Object Recognition using Radar Point Cloud Python, CV, Tensorflow Aug 2020 - Dec 2020

Implemented and performed experiments on the accuracy of VoxelNet and Point Net architectures on the radar point cloud.

Changed the model architectures to make it suitable for the sparse radar data. Combined the image feature nets to make the result more accurate.

Built a pipeline, starting with taking input from react-native mobile application, triggering cloud functions in GCP to do the required bounding box detection and send the output back to the mobile application. Doctor Consultancy System Python, Django, JavaScript Course : Applied Software Engineering Aug 2018 - Dec 2018

Partnered with 3 other classmates to create a full client and server implementation facilitating doctors to open time slots and for patients to book them.

Wrote algorithm for slot opening and booking. Worked on front-end for the doctor and patient home screen and dashboards.

Created a chat service option using web sockets. Technical Skills

Languages: Python, JavaScript

Frameworks and Databases: React, Node, Angular, Pytorch, Express, MySQL Libraries: HuggingFace, Pandas, NumPy, Matplotlib, Scikit-learn Tools: Git, Linux,Kubernetes, Docker Visualization Tools: Power BI

Contact this candidate