Date: May 2024 - Aug 2024
Roles: AI Engineer, Data Scientist
This project was completed while I was a consultant at Melio AI, and some of the project details have been obfuscated.
Overview
Developed multiple low-cost agentic RAG chatbot systems as proof of concept to demonstrate value for future development. Designed and built a reusable system template that was adopted company-wide for agentic LLM projects.
My Responsibilities
- Designed architecture for agents, knowledge bases, and interfaces.
- Built low-cost serverless knowledge bases and agent-agnostic APIs for quick experimentation.
- Instrumented interfaces for trace collection using Langfuse.
- Created evaluation pipelines using LLM-assisted testing for performance measurement.
- Deployed infrastructure with Terraform and tested using custom interfaces.
Outcomes/Impact
- Created a reusable system template that was adopted as a company-wide standard for agentic LLM pojects.
- Demonstrated chatbot value to clients for specific use cases.
Tools Used
- Python
- AWS
- AWS Bedrock
- AWS Bedrock Agents
- AWS Lambda
- LangChain
- LangServe
- Langfuse
- Anthropic Claude
- HuggingFace
- ChromaDB
- Gradio
- Terraform