Garvit Galgat

2024 · DATA QUOTIENT

AnalytiQ

From a dataset to analysis and machine-learning models

Python · Streamlit · Polars · H2O AutoML · OpenAI

I worked on AnalytiQ, a Streamlit application that brings data preparation, analysis, and machine learning into one place. Users can upload a dataset, clean it up, explore it, and compare models in the same app.

Understand and prepare the data

Users can manage dataset versions, check for missing or duplicate values, and apply data-quality rules. Charts help users explore distributions and relationships between columns before preparing the data for a model.

Compare models with context

Users can train classification and regression models with H2O AutoML and compare their results on a leaderboard. An LLM helps suggest algorithms and which column to predict based on the use case, then explains the results. The user chooses what to train.