A quick overview of my background and skills...
I’m skilled at bridging the gap between technical and non-technical audiences so decisions are clear and aligned.
AI-powered customer chatbot for a home improvement company using an LLM agent that directs users to the right RAG system for either kitchen remodeling advice or store policy questions. The agent decides which RAG tool is best-suited to answer the query.
This is a multi-tool agent framework built with LangChain and OpenAI’s LLM. The agent chooses between four integrated tools: Wikipedia for factual recall, llm-math for computation, ddg-search for real-time web data, and arXiv for scholarly research.
The Hungry Bot Fast Food Chatbot is a speech-enabled ordering system that combines speech-to-text (STT), text-to-speech (TTS), and rule-based logic to handle customer orders. It listens to a customer's order, applies rules, and confirms the order verbally.
The AWS Support Triage Agent, built with Amazon Bedrock, intelligently routes customer inquiries to the right support domain. It classifies questions — such as product troubleshooting, billing, or policy — and retrieves relevant information from an S3 knowledge base.
The AWS News Digest Summarizer automatically gathers top headlines from multiple RSS feeds each morning, summarizes them using Bedrock, and stores the results in S3. The results are emailed using SNS, creating a fully serverless daily news pipeline.
HealthScan AI is a multimodal voice-and-vision system built on AWS that listens to a user’s question, analyzes a skin image using a SageMaker model, and speaks back an explanation with confidence scores — showcasing intelligent audio, visual, and language integration.
Please reach out, I'd like to connect!