Calliv

The challenge

Scientific data rarely arrives analysis-ready. Calliv set out to take raw scientific datasets and make them clean, processable, and legible — turning messy CSV inputs into something a human can actually reason about and visualize.

My role

Built the project end to end: the data-processing pipeline and the server-rendered interface that exposes it.

Approach

• Data cleanup & processing — handled the unglamorous but essential work of normalizing and validating raw CSV input before anything downstream could trust it.
• Front end (Vue + Nuxt, SSR) — used server-side rendering for fast first loads and a UI geared toward presenting processed data rather than decorating it.
• Analysis foundation (TensorFlow) — set up the groundwork for applying machine-learning techniques to the cleaned datasets.

What it demonstrates

• The data-science fundamentals that matter most in practice: cleaning and validation come before modeling.
• An ability to move between the data layer and a presentation layer that makes findings legible.
• Choosing the right rendering strategy (SSR) for the job rather than reaching for a default.

Source: github.com/iasisalomon/calliv