Nagavalli Mareedu

Developed and deployed LLM-based applications for chat assistants, document Q&A, and summarization use cases, improving knowledge access and response efficiency by 30%.
Designed and implemented RAG pipelines using LangChain, LlamaIndex, and FAISS to improve context-aware information retrieval accuracy by 35%.
Fine-tuned Large Language Models (LLMs) including GPT, LLaMA, and Claude on domain-specific datasets to improve response accuracy and relevance by 25%.
Built scalable Generative AI microservices using FastAPI, REST APIs, and GraphQL, containerized with Docker and orchestrated with Kubernetes.
Developed user-facing AI interfaces using React and Next.js for real-time interaction with Generative AI systems.
Implemented prompt engineering, token optimization, caching, and streaming techniques to improve inference performance and reduce latency.
Automated model deployment and monitoring workflows using CI/CD pipelines, GitHub Actions, OpenTelemetry, and Prometheus, reducing release cycles by 40%.
Collaborated with cross-functional teams to deliver production-grade AI solutions that improved automation, scalability, and user experience.

Developed and deployed LLM-based applications for chat assistants, document Q&A, and summarization use cases, improving knowledge access and response efficiency by 30%., Designed and implemented RAG pipelines using LangChain, LlamaIndex, and FAISS to improve context-aware information retrieval accuracy by 35%., Fine-tuned Large Language Models (LLMs) including GPT, LLaMA, and Claude on domain-specific datasets to improve response accuracy and relevance by 25%., Built scalable Generative AI microservices using FastAPI, REST APIs, and GraphQL, containerized with Docker and orchestrated with Kubernetes., Developed user-facing AI interfaces using React and Next.js for real-time interaction with Generative AI systems., Implemented prompt engineering, token optimization, caching, and streaming techniques to improve inference performance and reduce latency., Automated model deployment and monitoring workflows using CI/CD pipelines, GitHub Actions, OpenTelemetry, and Prometheus, reducing release cycles by 40%., Collaborated with cross-functional teams to deliver production-grade AI solutions that improved automation, scalability, and user experience.

Designed and deployed a Generative AI-powered Q&A system enabling users to query unstructured enterprise documents through natural language.
Built RAG pipelines using LangChain, LlamaIndex, and vector databases, improving context-aware document retrieval accuracy by 35%.
Integrated OpenAI GPT-4 with custom prompt templates and dynamic context injection to generate grounded and relevant responses.
Developed backend APIs using FastAPI, containerized with Docker, and deployed on AWS ECS with S3 storage and CloudWatch monitoring.
Fine-tuned sentence-transformer models and implemented chunking/indexing strategies to improve semantic search quality by 30%.
Automated embedding refresh and vector index updates using PostgreSQL triggers, improving data freshness and retrieval reliability.
Monitored system performance using LangSmith, Prometheus, CloudWatch, and internal logging to improve auditability and maintainability.
Evaluated LLM providers including OpenAI, Cohere, and Anthropic based on accuracy, latency, cost, and scalability.
Implemented guardrails, fallback logic, response validation, and structured outputs using OpenAI function calling.

Designed and deployed a Generative AI-powered Q&A system enabling users to query unstructured enterprise documents through natural language., Built RAG pipelines using LangChain, LlamaIndex, and vector databases, improving context-aware document retrieval accuracy by 35%., Integrated OpenAI GPT-4 with custom prompt templates and dynamic context injection to generate grounded and relevant responses., Developed backend APIs using FastAPI, containerized with Docker, and deployed on AWS ECS with S3 storage and CloudWatch monitoring., Fine-tuned sentence-transformer models and implemented chunking/indexing strategies to improve semantic search quality by 30%., Automated embedding refresh and vector index updates using PostgreSQL triggers, improving data freshness and retrieval reliability., Monitored system performance using LangSmith, Prometheus, CloudWatch, and internal logging to improve auditability and maintainability., Evaluated LLM providers including OpenAI, Cohere, and Anthropic based on accuracy, latency, cost, and scalability., Implemented guardrails, fallback logic, response validation, and structured outputs using OpenAI function calling.

Supported Data Science and Machine Learning coursework by guiding 100+ students on Python, AI concepts, and model-building assignments.
Assisted in developing AI/ML course materials, lab exercises, and learning resources, improving student engagement by 30%.
Facilitated workshops and lab sessions explaining Machine Learning, NLP, and AI concepts to technical and non-technical students.
Helped integrate AI tools and Moodle LMS platforms into academic workflows, improving learning accessibility by 20%.
Guided students in completing AI-assisted assignments while maintaining academic integrity and responsible AI practices.
Supported development of guidelines for the ethical and responsible use of AI in academic environments.

Supported Data Science and Machine Learning coursework by guiding 100+ students on Python, AI concepts, and model-building assignments., Assisted in developing AI/ML course materials, lab exercises, and learning resources, improving student engagement by 30%., Facilitated workshops and lab sessions explaining Machine Learning, NLP, and AI concepts to technical and non-technical students., Helped integrate AI tools and Moodle LMS platforms into academic workflows, improving learning accessibility by 20%., Guided students in completing AI-assisted assignments while maintaining academic integrity and responsible AI practices., Supported development of guidelines for the ethical and responsible use of AI in academic environments.

Designed and deployed Machine Learning models for classification, recommendation, and NLP tasks using Python, Scikit-learn, XGBoost, and PyTorch, improving prediction accuracy by 25%.
Built backend APIs and ML services using Python, Docker, and Kubernetes on Azure for scalable model deployment.
Developed end-to-end ML pipelines covering data ingestion, preprocessing, model training, validation, and deployment.
Integrated ML models into full-stack applications using REST APIs, Celery, Redis, React, Next.js, and Tailwind CSS.
Optimized data pipelines using PostgreSQL, MongoDB, and BigQuery, improving model input processing efficiency by 30%.
Managed model lifecycle with MLflow and DVC for experiment tracking, versioning, and reproducibility.
Deployed production models using CI/CD pipelines, GitHub Actions, Prometheus, and Grafana, reducing deployment time by 35%.
Built Streamlit and Plotly dashboards for model monitoring, A/B testing, and stakeholder reporting.

Designed and deployed Machine Learning models for classification, recommendation, and NLP tasks using Python, Scikit-learn, XGBoost, and PyTorch, improving prediction accuracy by 25%., Built backend APIs and ML services using Python, Docker, and Kubernetes on Azure for scalable model deployment., Developed end-to-end ML pipelines covering data ingestion, preprocessing, model training, validation, and deployment., Integrated ML models into full-stack applications using REST APIs, Celery, Redis, React, Next.js, and Tailwind CSS., Optimized data pipelines using PostgreSQL, MongoDB, and BigQuery, improving model input processing efficiency by 30%., Managed model lifecycle with MLflow and DVC for experiment tracking, versioning, and reproducibility., Deployed production models using CI/CD pipelines, GitHub Actions, Prometheus, and Grafana, reducing deployment time by 35%., Built Streamlit and Plotly dashboards for model monitoring, A/B testing, and stakeholder reporting.