MiMoHub: Conversational Excel & CSV Analysis

A business user with a spreadsheet needs a practical way to inspect the data and ask useful follow-up questions. MiMoHub brings file exploration, data-quality findings and conversational analysis into a single focused application.

My role: Python/Streamlit application development and AI integration. I built an Excel and CSV analysis interface with Pandas-based processing, Plotly visualizations and a Xiaomi MiMo connection through LiteLLM.

Upload a spreadsheet, inspect summary statistics and distributions, review missing values, duplicates and outliers, then ask questions in plain language. The workflow combines deterministic data checks with model-assisted exploration so the user can examine the dataset before interpreting an answer.

The deliverable connects a user-facing interface, data-processing logic and an external model in one application. It demonstrates the pattern behind internal analytics assistants: a defined input, useful checks, an understandable interface and a conversational layer.

Independent public build. The linked repository contains the application source, processing utilities, dependencies and local setup instructions. Model-powered chat requires a configured provider key. A client version would be scoped around its data, access rules and review requirements.

Want an assistant for a recurring spreadsheet workflow? Send a non-sensitive sample structure, the questions users ask and the output they need. I can scope a first milestone around one dataset and a testable end-to-end flow.