Skip to content
View phoenixha4's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report phoenixha4

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
phoenixha4/README.md

Harsh Agarwal · AI/ML Engineer

Designing and shipping multi-agent, multi-turn conversational pipelines for real-world banking & financial workflows. Deeply interested in the infrastructure layer of agentic AI — memory, observability, context engineering, and evals.

Frameworks & Orchestration LangGraph AutoGen FastAPI Docker Azure

Models & Tooling Opik HuggingFace Chainlit


🔭 Currently Exploring

AG-UI protocol · MCP · AI evals · Memory architecture · Voice AI agents


📌 Notable Projects

Project Description
GenerativeUI Agentic onboarding with LangGraph + Chainlit — backend drives dynamic form generation via structured LLM outputs
AgenticClassification Multi-step text analysis pipeline with LangGraph
LearnDocker Sentiment analysis model → FastAPI → Docker
what2eat Vegan recipe recommender system

LinkedIn Email

Pinned Loading

  1. DynamicUI DynamicUI Public

    An AI-powered agent application using LangGraph and Chainlit for conversation-based journeys with an intelligent and dynamic interface.

    Python

  2. AgenticClassification AgenticClassification Public

    Demonstrates the power of LangGraph by building a multi-step text analysis pipeline

    Jupyter Notebook

  3. LearnDocker LearnDocker Public

    This project demonstrates how to wrap a machine learning model in a FastAPI web application and Dockerize it for easy deployment. The application performs sentiment analysis using Hugging Face's tr…

    Dockerfile

  4. slate slate Public

    Slate Todo is a Go-based full-stack todo app with PostgreSQL, Docker, and a vanilla JS frontend, built to learn backend structure, APIs, and deployment in Go.

    Go

  5. what2eat what2eat Public

    This project is a Vegan Recipe Recommender system that provides recipe recommendations based on user preferences and/or input ingredients.

    Jupyter Notebook 5