Section A
My Primary Contributions
End-to-End AI Engineering
Architected and deployed the platform's predictive AI microservice from the ground up.
Data Pipeline & Preprocessing
Developed robust data pipelines using Pandas. Cleaned raw historical data, parsed complex `budgetallocation` strings, and engineered custom analytical features such as `budget_per_head`.
Predictive Modeling
Trained an ensemble RandomForestClassifier for intelligent venue recommendation and a multi-output RandomForestRegressor for granular, percentage-based budget allocations.
Interactive Dashboard
Wrapped the predictive models into a fully interactive Streamlit application that dynamically calculates event costs and visualizes budget distributions in real-time.
Section B
Overall Project Capabilities
Working in a cross-functional Agile team, the platform also featured the following full-stack integrations:
Full-Stack Ecosystem
A modern, highly responsive frontend built with Next.js (TypeScript) connected to a robust Node.js/Express RESTful API.
Database Architecture
Complex, ACID-compliant database schemas managed in PostgreSQL to handle users, event parameters, and secure transactions.
Cloud Infrastructure
Utilization of AWS S3 for decoupled document management and asset storage.