Software Engineer
Summary
Working on VoiceX platform which is used to create AI voice agents out of the box. Building a SDK to evaluate the agents on the metrics given by the user, designed the architecture for the SDK which adapts to any external eval platform for evaluating the agents and also adapts to in-house agent platform.
Highlights
Technologies Used : Python, FastAPI, AWS, OpenAI API, langchain, PostgreSQL
Software Engineer
Summary
Designed and implemented a highly efficient relational database schema in MySQL for an application that supports seamless interaction with multiple large language models (LLMs). Used RAG(Retrieval-Augmented Generation) over the chat history embeddings that are stored in a vector database to search for a closest chunk and pass as a context along with the user prompt for accurate responses. Implemented Knowledge Directories which is a collection of embeddings stored in Qdrant database of the files uploaded by the user which are used when user wants response for the query using the knowledge of directory. Optimized database structure for fast retrieval and storage of chat history, ensuring scalability and performance. Created normalized data models to reduce redundancy and improve data integrity across various application components. Applied best practices for normalization (up to 3NF) to ensure consistency and data quality. Created services in Python using FastAPI, SQLAlchemy and MySQL, for user authentication using OAUth2, for storing and retrieval of chat history and sessions, for fetching the available models based on multi-tenant policy. Created user service for file upload and retrieval using pre-signed URL from AWS S3 bucket Optimized the architecture of a service which fetches the images which reduced the AWS Infra cost by 20%. Created notification services using Spring Boot, AWS SNS, and Firebase.
Highlights
Technologies Used: Flutter, FastAPI, Qdrant, SQLAlchemy, Uvicorn, Python, AWS, MySQL, Spring Boot.
Bachelors
Computer Science
Grade: 7.55/10.0
Courses
Data Structures
Design and Analysis of Algorithms
Operating Systems
Computer Networks
Machine Learning
DBMS
Paralle Computing(Openmp, Cuda)
Distributed Systems
C, C++, Dart, Python, Java, Javascript.
Flutter, HTML, CSS, Javascript, React.js, jQuery.
FastAPI, Spring Boot, Express.js.
RAG, AI Agents, Ollama, vector database, langchain, OpenAI, Qdrant.
MySQL, MongoDB.
Data Structures, Algorithms, Git, Redis.
Summary
This is an AI Agent framework used to create Agents, which is developer friendly. The main goal of this project is to simplify the understanding of the flow of the messages that are exchanged with the LLM.Showing the events that are involved like function execution, passing the result of the executed function, etc.