Machine Learning Engineering
Ranking Top 50 Industries on Industry Level from NACIS Data Using Python, SQL, Airtable- Extracted and cleaned industry data, performing feature selection using Python libraries such as Pandas, Matplotlib, NumPy, and SciPy.- Applied principal component analysis (PCA) on logistic regression, decision trees, and CNN models using Scikit-learn, which ranked the top 50 industries by weighting feature importances and scores across different data levels.AI Agent Framework and Tool Development using LangChain, LangGraph and Large Language Model (LLM)- Developed LLM agent models to implement RAG models in Python using LangChain and LangGraph, enabling self-reflection on intermediate outputs, and regeneration of correct JSON format before passing to the next stage.- Defined and developed 3 customized tools for LLM agents on generating metadata JSON, which validated required and optional parameters in API calls, and detected missing placeholders in dynamic inputs for seamless model execution.
