Lee Anwar

Owned end-to-end development of AI agents using LangChain and React prompting to automate complex workflows, aligning with customer-focused solutions and innovative simplification. Scaled RAG-based knowledge systems using AWS OpenSearch and vector databases to improve LLM accuracy for enterprise applications, similar to Amazon Kendra and Alexa. Implemented evaluation frameworks to ensure production-ready AI systems. Delivered optimized LLMs by fine-tuning open-source models on proprietary datasets. Deployed serverless AI solutions on AWS with CI/CD pipelines for seamless updates. Developed prompt engineering tools and dataset versioning systems to streamline experimentation. Communicated technical strategy through whitepapers and architecture reviews to promote scalable AI adoption.

Owned end-to-end development of AI agents using LangChain and React prompting to automate complex workflows, aligning with customer-focused solutions and innovative simplification. Scaled RAG-based knowledge systems using AWS OpenSearch and vector databases (FAISS, Pinecone) to improve LLM accuracy for enterprise applications, similar to Amazon Kendra and Alexa. Drove model excellence by implementing evaluation frameworks (BLEU, ROUGE, hallucination detection) to ensure reliable, production-ready AI systems. Delivered optimized LLMs by fine-tuning open-source models (LLaMA, GPT-J) on proprietary datasets, focusing on deep customer needs. Deployed serverless AI solutions on AWS (Lambda, ECS, SageMaker) with CI/CD pipelines for seamless, zero-downtime updates. Accelerated team productivity by developing prompt engineering tools and dataset versioning systems to streamline experimentation. Communicated technical strategy to leadership and partners through whitepapers and architecture revi

Designed and trained custom large language models using Hugging Face, BERT, RoBERTa, and GPT-2 for specialized applications including legal document processing, customer service automation, and technical content generation. Developed and deployed enterprise-grade NLP systems for intelligent search, document summarization, and conversational AI, enhancing operational efficiency and user experience. Managed the complete ML pipeline from data collection to deployment and performance monitoring. Created evaluation frameworks for ethical considerations including bias detection and factual accuracy verification. Established MLOps workflows with model versioning, automated retraining, and A/B testing. Conducted technical training sessions on NLP and mentored engineering teams.