Powell Molleti

Building a 0→1 multi-agent framework integrating open and closed models to automate marketing workflows and designing data and prompt pipelines to provide trustworthy context and telemetry for production agents.

Developed a Python SDK and backend multi-agent framework using LLaMA 3.1 inference, delivered MCP end-to-end integrations for enterprise tools, and exposed RAG data and actions as audited, schema-validated resources for production use by 1000+ customers.

Architected and deployed AWS data pipelines supporting RAG-based chatbots for MBA courses, evaluated Azure OpenAI vs Claude for complex table and image extraction, and used DeepEval to validate large-document summarization.

Led engineering team to accelerate legal discovery by deploying fine-tuned LLMs (Falcon 40b, LLaMA v2 70b) with QLoRA and SFT, architected RAG pipelines for long-context legal documents while ensuring data privacy and audited execution.

Developed conversational AI solutions for customer service using PyTorch on AWS SageMaker and designed data acquisition strategies to improve model performance with limited labeled data.

Consulted on NLP and deep learning projects across clients, leading clinical-trial NLP at Johnson & Johnson, architecting chatbots on GCP Vertex AI for PwC, and deploying face identity tracking pipelines for Verizon at 30+ fps.

Built deep learning infrastructure on AWS SageMaker with custom Docker containers, and managed continuous dataset creation using Spark on S3 to support ML/DL teams in production.

Developed a PySpark streaming pipeline ingesting Kafka streams and scaled it to handle multiple data sources on AWS EMR for reliable real-time processing.

Led design and development of scalable big data pipelines using Apache Spark and Flink to process over 50 million events per minute, enabling near-real-time analytics for enterprise security.

Led design and development of scalable big-data pipelines using Apache Spark and Flink, processing over 50 million events per minute to enable near-real-time analytics for enterprise security.

Completed a Graduate Certificate in Artificial Intelligence covering core AI concepts and applied machine learning techniques.

Completed a Graduate Certificate in Artificial Intelligence with coursework covering machine learning and AI applications.

Earned a Bachelor of Technology degree in engineering with foundational coursework in computer science and engineering principles.