Uday Kiran

About Uday Kiran

I am a Generative AI Engineer / AI Engineer with 6 years of experience building production-ready AI systems across LLMs, Retrieval-Augmented Generation (RAG), and Applied AI/ML at enterprise scale. I have designed and deployed GPT-4–powered solutions using LangChain and AWS Bedrock to deliver document intelligence, conversational AI, and decision-support platforms, with a strong focus on prompt engineering, embeddings, semantic retrieval, and MCP-aligned context management. My work includes developing Dockerized, AWS-deployed AI services backed by PostgreSQL/MySQL, and implementing end-to-end MLOps pipelines covering CI/CD, versioning, monitoring, and rollback to ensure reliable production releases.Alongside Generative AI, I bring a solid Applied AI foundation, having built large-scale AI systems using the Hadoop ecosystem with Spark/PySpark, and trained and operationalized supervised and unsupervised models using TensorFlow, Scikit-Learn, Spark MLlib, and XGBoost. I have owned the full model lifecycle from feature engineering and distributed data processing to evaluation, monitoring, and retraining. I enjoy collaborating with cross-functional teams to translate real-world business problems into scalable, cloud-native AI solutions, and I’m particularly interested in roles focused on GenAI platforms, LLM-driven applications, RAG architectures, and AI systems at scale.

Experience

1. Generative Ai Engineer

   Cintas

   Mar 2023 — Present · Mason, OH, US

   Designed and delivered enterprise Generative AI solutions supporting LLM inference and Retrieval-Augmented Generation (RAG) workflows for analytics and decision-support systems. Built LLM-powered applications for text generation, document summarization, and insight extraction using LangChain, prompt engineering, and GPT-4 via OpenAI / Azure OpenAI, with inference routed through AWS Bedrock. Established end-to-end MLOps pipelines for GenAI, covering prompt versioning, model configuration management, automated deployment, and rollback across environments. Developed RAG pipelines leveraging embeddings, semantic context optimization, and vector search to ground LLM responses in enterprise data. Containerized GenAI services using Docker and deployed them on AWS compute platforms (ECS / EKS-style architectures) for scalable and reliable execution. Designed and maintained vector-based retrieval systems, persisting prompt metadata, conversation state, embedding references, and audit logs in PostgreSQL / MySQL. Developed RESTful APIs using Python, FastAPI, and Flask to orchestrate LLM calls, manage embeddings, and persist AI-generated outputs in relational databases. Implemented CI/CD pipelines for GenAI systems, automating Docker builds, deployments, database migrations, versioning, monitoring, and lifecycle management of LLM and RAG workflows.