Machine Learning Artificial Intelligence

Vigneswar Sundaramurthy

908-***-**** Philadelphia, PA *********.******@*****.*** Linkedin Portfolio EDUCATION

Drexel University Philadelphia, PA

Master of Science in Computer Engineering September 2022 - June 2024 Honors: Magna Cum Laude

Post Graduation program in Machine learning and Artificial intelligence The University of Texas Austin, Online Masters in Artificial Intelligence and Machine learning March 2020 - September 2021 Bachelor of Technologyin Electronics and Communications Engineering Kerala Technical University Bachelors in Electronics and Communications Engineering June 2015 - December 2019 SKILLS

• Frameworks/ Platforms: Amazon Web Services( AWS), Microsoft Azure, Google Cloud Platform (GCP), Vertex AI, Amazon Sagemaker, LangChain, LlamaIndex, TensorFlow, PyTorch, Caffe, JAX, Keras, OpenCV, Scikit-learn, Windows, Linux, Jupyter, JupyterLab, Visual Studio Code, Databricks, Neo4j, Docker, Kubernetes, Tesseract, Pandas, Numpy, Scikit, scipy, xgboost

• Specializations: Artificial Intelligence, Deep Learning, Statistical Modelling, Data Analytics, Natural Language Processing, Recommendation Systems, Graph Databases, Vector Databases(Pinecone, Elasticsearch), Web Scraping, APIs, Prompt Engineering, Reinforcement Learning, Computer Vision

• Programming Languages: Python, C, C++, MATLAB, PySpark, SQL, R, JavaScript, HTML, CSS, Node.js, Go, HTML, CSS, React CERTIFICATIONS

• Microsoft Certified AI Engineer Associate September 2024 Validates expertise in using Microsoft Azure AI to build and manage AI solutions that leverage machine learning and cognitive services.

• Databricks Accredited Generative AI Fundamentals October 2024 Validates expertise in using Databricks to build and manage Gen AI solutions that leverage machine learning and cognitive services.

• PCEP – Certified Entry-Level Python Programmer November 2024 Certified Python Programmer accredited by the Python Institute, demonstrating expertise in Python programming and a commitment to profes- sional development.

WORK EXPERIENCE

Prompt Owl Atlanta, GA

AI Engineer Present

• Developed a Retrieval-Augmented Generation (RAG) system using a multi-query approach, improving response accuracy by 20% through fusing multiple responses from Large Language Models (LLMs).

• Created a scalable document repository that processes and makes over 100,000 documents queryable in real time, ensuring seamless and precise information retrieval.

• Implemented a web scraping feature that allows users to input URLs, scrape data from 50+ websites, and query the database, enriching information with over 10,000 new data points.

• Improved data accessibility by integrating LLMs with web-scraped data, increasing query resolution time by 30% and enhancing user engage- ment.

• Deployed the application using AWS Lambda, enabling serverless architecture for efficient, scalable, and cost-effective operations.

• Deployed the application in a Docker container on a private server running at Private Server Deployment, enabling robust, scalable, and cost- effective operations.

Tools: LangGraph, LangChain, LangSmith, Neo4j, CrewAI, KnowlegeGraph, LlamaIndex, Python, BeautifulSoup, GPT-4, TensorFlow, Neo4j, Elasticsearch, Docker, Google Cloud Platform (GCP), Streamlit(for deployment) Zoetis Parsippany, NJ

Data Science Intern June 2023 - September 2023

• Engineered and delivered a Virtual sensor for Zoetis, enabling precise measurement of Lactate Levels in Bio Reactors project completed ahead of schedule, within a 12-week timeframe, ensuring timely product launch and market advantage

• Implemented machine learning techniques, achieving a 99% predictive accuracy using Random Forests, SVM, CNN, and Gradient Boost- ing.These optimizations significantly enhanced real-time cell health monitoring

• Synchronized and analyzed unstructured data from 20+ sources, resulting in a 15% increase in cost-effectiveness, saving approximately

$25,000. Managed insights from over 1,000,000 data points Tools: Databricks, Data Analysis, Statistical Modelling, Feature Engineering, Git (Version Control), Data Integration, SQL Server Management Studio, Azure Blob Storage, Azure SQL Database, Shiny, Posit Tata Consultancy Services Bangalore, India

Systems Engineer January 2020 - September 2022

• Streamlined defect detection & analysis through a machine learning model, reducing issues 25% in half a year as per SCRUM Agile Project Management

• Diagnosed and optimized In-vehicle networks (CAN/LIN/Ethernet) for rigorous physical layer tests for seamless operation and compliance

• Integrated machine learning into CAN testing slashed test time by 30%, boosted coverage by 15%, and reduced manual errors by 20%, saving

$50,000 annually

Tools : TensorFlow, MATLAB, Python,SQL, PySpark (data processing), C PROJECTS

Patent NLP Philadelphia, PA

Project Link: https://github.com/vigneswar96/Patent_NLP November 2023 - July 2024

• Designed and implemented a sophisticated information retrieval system using semantic search techniques, improving patent similarity matching by 40% compared to traditional keyword-based methods.

• Designed and implemented a multi-input transformer-based model leveraging BERT embeddings for NLP tasks, integrating multiple text sources through concatenation and pooling. Achieved efficient feature extraction using GlobalAveragePooling and optimized performance with dense layers and dropout, managing 918,529 trainable parameters for improved prediction accuracy.

• Utilized transfer learning techniques to fine-tune pre-trained language models on a corpus of over 1 million patent documents, achieving a 95% accuracy in patent classification tasks.

Tools: BERT, PyTorch, Hugging Face Transformers, spaCy, NLTK, Elasticsearch, Flask, Docker, Google Cloud Platform(GCP) Reinforcement Learning: Training Atari Game Players with TensorFlow Agents Philadelphia, PA Project Link: https://github.com/vigneswar96/appliedML/tree/main/Tensorflow_agents/Notebooks April 2023 - May 2023

• Optimized Atari games using TensorFlow agents, achieving peak performance with the highest score

• Executed diverse game levels utilizing TF-Agent, encompassing skill levels from Poor to Intermediate and High, and systematically preserved the outcomes

Tools: Q-learning, SARSA, Deep Q Networks, Policy Gradient Methods, OpenAI Gym, RLlib, Google Cloud Platform (GCP) Integrating Vision and Generative Models Philadelphia, PA Project Link: https://github.com/vigneswar96/appliedML/tree/main/Generative_AI/Notebooks March 2023 - April 2023

• Instituted a Real-time image processing with Stable diffusion for seamless subject replacement, ensuring precision and efficiency

• Developed custom datasets by curating and annotating thousands of images for real-time object detection and subject replacement, ensuring high-quality training data for YOLOv5 and YOLOv8 models.

• Implemented data augmentation techniques such as rotation, scaling, and flipping to increase dataset diversity and improve model generaliza- tion, resulting in more robust image recognition performance.

• Utilized segmentation tools like Segment Anything Model (SAM) to automate object labeling, significantly reducing manual effort and en- hancing dataset accuracy by 30%.

Tools: YOLOv5, YOLOv8, Segment Anything Model (SAM), PyTorch, OpenCV AI Powered content retrieval system Philadelphia, PA Project Link: https://github.com/vigneswar96/AI-Powered-Content-Retrieval- June 2024 - August 2024

• Implemented a AI Powered Content Retrieval system. This could convert all the documents loaded into a vector database and then could be queried for analysis.

• Developed a cutting-edge Retrieval-Augmented Generation (RAG) system ( using Langchain) that transforms unstructured documents into high-dimensional vectors, enabling efficient, high-precision information retrieval across large document collections. Tools: Langchain, Ollama, Llama 3.3, Groq, Linux, Langsmith ( Tracing) Research Publications

• Published IEEE Journal Paper titled: “Machine Learning Approach for Clustering of Countries to identify the best Strategies to combat Covid 19”. Presented in IEEE Toronto 2021 International Conference Project Link: https://ieeexplore.ieee.org/abstract/document/9422621 September 2020 - November 2020

• Published “Detecting Key Soccer match events to create highlights using Computer Vision” in arxiv Project Link: https://arxiv.org/abs/2204.02573 June 2021 - September 2021

Contact this candidate