Kamesh Sadhasivam

COGNIZANT TECHNOLOGY SOLUTIONS, Chennai, India Programmer Analyst (Artificial Intelligence & Analytics)  September 2023 – December 2025 Designed, trained, and deployed machine learning models (CNNs, anomaly detection) to identify patterns and irregularities in large datasets, achieving 95% accuracy and reducing false positives by 20%.        Built and maintained automated data pipelines for data collection, cleaning, transformation, and processing using Python and SǪL. Performed data wrangling, validation, and quality checks to ensure accuracy and suitability for analytics and modelling. Developed predictive and classification models to analyze trends, forecast outcomes, and identify operational risks and opportunities. Implemented statistical and machine learning techniques using Scikit-learn and TensorFlow. Collaborated with cross-functional teams and business stakeholders to translate analytical results into actionable insights. Developed and deployed RESTful APIs to integrate ML and GenAI models into enterprise applications. Worked on Generative AI implementations across the SDLC, improving automation and operational efficiency. COGNIZANT TECHNOLOGY SOLUTIONS, Chennai, India AI/ML Intern  May 2023 – September 2023 Automated testing and data-driven workflows using Python and Selenium, improving efficiency by 30%.  Applied data structures, algorithms, and analytical problem-solving to optimize system performance. CERTIFICATIONS and WORKSHOPS • • • • • Microsoft Azure Fundamentals Cloud architecture, Azure Services and Management tools on Azure Oracle Academy Foundation of Database Management System Concepts FICE Education Private Limited Artificial Intelligence and High-Performance Computing Association of Mechatronics Engineers Workshop on Sensory Perception for Autonomous Driving Using Python NVIDIA Deep Learning Institute Building Video AI Application using Jetson Nano PROJECTS Title: REAL-TIME VIDEO ANALYTICS ON NVIDIA JETSON NANO Summary: Deployed a YOLOv5-based object detection model on Jetson Nano using TensorRT for hardware acceleration. Optimized the pipeline with Deep Stream SDK, achieving 22 FPS at 720p. Integrated MQTT for edge-to-cloud alerts and reduced power consumption by 30% via motion-triggered inference. Tools: Python, OpenCV, TensorRT, ONNX. Title: KNEE-OSTEOARTHRITIS SEVERITY PREDICTION Using Deep Learning Summary: Developed a deep learning-based application to classify knee osteoarthritis severity (Healthy, Moderate, Severe) from X-ray images. Implemented CNN architectures (Manual CNN, VGG-16, LeNet-5), achieving up to 99% accuracy with LeNet-5. Preprocessed and augmented 4,500+ images using rescaling, histogram equalization, zoom, and flipping to enhance model generalization. Deployed the trained model via Django with an HTML/CSS interface for real-time severity prediction. Compared models based on accuracy, recall, precision, and misclassification percentage to determine optimal performance. Demonstrated the effectiveness of deep learning in medical imaging for early, automated diagnosis, enabling faster clinical decision-making. Title: GENERATIVE-AI TEST CASE GENERATOR Summary: Developed an AI-driven system to generate positive and negative test cases from unstructured requirement documents using LLMs (GPT-4, LLaMA-2) and Lang Chain. Implemented OCR and NLP pipelines for requirement extraction, stored embeddings in Pinecone for semantic retrieval, and automated structured test case creation with validation. Enabled export to Excel/CSV/JSON and integration with Jira/TestRail/Azure DevOps, reducing manual QA effort by 75%. Title: GEN-AI POWERED KNOWLEDGE TRANSFER ASSISTANT for Seamless Knowledge Management Summary: Engineered an AI-driven onboarding assistant leveraging GPT-4/LLaMA-2 via Lang Chain to transform unstructured Knowledge Transfer assets (PDFs, docs, videos) into interactive, searchable knowledge. Built NLP pipelines with OCR and speech-to-text, storing vector embeddings in Milvus/Pinecone for semantic retrieval and delivered contextual Q&A and keyword search across multimodal data, with exportable results (PDF/Excel/JSON) and dashboard integration—cutting onboarding time and senior SME reliance by 60%. TECHNICAL SKILLS Programming & Scripting: Python, SQL, C++ Machine Learning: Scikit-learn, TensorFlow, CNNs, Predictive Modelling, Anomaly Detection Data Engineering: Data Pipelines, ETL, Data Cleaning, Data Transformation Gen-AI & NLP: GPT-4, LLaMA-2, Lang Chain, Vector Databases Cloud & DevOps: Azure, Docker, Kubernetes, CI/CD, Git, Bitbucket Databases: dB Visualizer, SQL Databases, Power Apps ROLES and RESPONSIBILITIES • • Represented the Gen-C community at the Cognizant Delivery Summit-2024, highlighting my project with PepsiCo Frito-Lay North America. Served as a Gen-C representative at Cognizant's North America client year-end celebration. • Organized Events and workshops for Cognizant’s client visit.