Vineeth Reddy kothakapu

AI/ML Engineer with 4.5 years of experience designing and deploying scalable Machine Learning, Deep Learning, Generative AI, and LLM-powered applications across enterprise environments. Experienced in building end-to-end AI solutions, including data pipelines, model development, RAG architectures, multimodal AI, agentic workflows, MLOps, and cloud-native deployments. Proven ability to deliver production-ready AI systems that improve model performance, scalability, and business outcomes through data-driven innovation.

ML Engineer - Cigna Healthcare - USA

(2025-03)

• Architected enterprise-scale RAG solutions processing 5M+ healthcare records, improving contextual search relevance by 35% using hybrid retrieval (BM25 + vector search).
• Built end-to-end LLM pipelines integrating retrieval, prompt engineering, and response generation to automate clinical documentation and medical text summarization.
• Optimized vector search using FAISS and Pinecone, reducing query latency by 25% while improving semantic search accuracy.
• Developed and optimized ML/DL models (CNNs, RNNs, Transformers) for predictive analytics and healthcare classification, improving forecasting accuracy by 28% and model precision by 25%.
• Evaluated BERT and GPT models, balancing accuracy, latency, and cost to deploy production-ready AI solutions.
• Developed multi-agent AI workflows using LangGraph, implementing tool calling, memory management, and reasoning for intelligent workflow automation.
• Implemented LLM evaluation frameworks and prompt optimization techniques to improve response quality, reduce hallucinations, and enhance production model reliability.
• Built scalable FastAPI services, REST APIs, and microservices to deploy AI models, enabling secure, low-latency real-time inference.
• Improved inference performance through batching, caching, and asynchronous processing, increasing throughput while reducing API response costs.
• Designed scalable cloud-native AI architectures using AWS services, enabling secure deployment, efficient resource utilization, and high-availability inference workloads.
• Implemented MLOps pipelines for automated deployment, model monitoring, versioning, and A/B testing, ensuring reliable and scalable AI operations.
• Collaborated with Product Managers, Data Scientists, Data Engineers, and healthcare stakeholders to deliver AI-powered solutions that enhanced operational efficiency and clinical decision-making.

Data Scientist/ML Engineer - Cognizant - India

(2022-12 - 2023-12)

• Constructed machine learning models (XGBoost, Random Forest) for customer behavior analysis and conversion prediction, increasing conversion rates by 22%.
• Transformed raw data through preprocessing and feature engineering, improving model performance and reliability.
• Designed NLP-based pipelines for text classification and information extraction from unstructured data.
• Developed transformer-based NLP solutions using BERT and LLMs for semantic search, document intelligence, and intelligent text automation, improving information extraction and search accuracy.
• Analyzed large-scale datasets to identify patterns, anomalies, and business drivers.
• Streamlined SQL queries and data pipelines to support efficient model training and deployment workflows.
• Visualized insights through dashboards, enabling data-driven decision-making.
• Automated reporting and ML workflows, reducing manual effort and improving efficiency.

Jr. Data Scientist - Tiger Analytics - India

(2021-01 - 2022-12)

• Developed fraud detection and anomaly detection models using supervised and unsupervised machine learning techniques, reducing fraudulent transactions by 18% across real-time processing environments.
• Built NLP pipelines for document classification and text processing by implementing tokenization, text preprocessing, and vectorization techniques, improving processing throughput by 35%.
• Implemented CNN-based deep learning models for image classification and object detection, achieving 92% validation accuracy across benchmark and production datasets.
• Performed exploratory data analysis (EDA) and feature engineering to identify high-impact variables, improving model precision and reducing data noise.
• Developed scalable data ingestion and preprocessing pipelines for structured and unstructured datasets, supporting efficient model training, validation, and testing workflows.
• Integrated machine learning models into production applications using REST APIs and scalable deployment pipelines, enabling secure, real-time predictions and improved operational efficiency.
• Optimized model performance through hyperparameter tuning and efficient data processing techniques, reducing computational costs by 25% while maintaining model accuracy.

Masters - Business Information Systems - Central Michigan University

B.Tech - Electronics and Electrical Engineering - St. Peters Engineering College