VizChitra 2025: India's first community-driven Data Visualization conference

In this live session, we’ll walk through how to go from a bare prompt to a polished, shareable chart—entirely by conversing with an LLM. Along the way, you’ll see real demos, such as this Web Features explorer, Elimination Game, LLM mental math evaluation, etc. and learn to:

• Use ChatGPT instead of Jupyter Notebooks, Figma, or D3 as the canvas for data storytelling
• Engineer prompts that generate valid data analyses and visualizations
• Iteratively refine axis scales, colors and annotations without writing a single line of code
• Automate interpretation: ask the model for captions, tooltips, and data insights
• Spotlight future trends: what generative data art looks like in six months

TAKEAWAYS

• A “prompt recipe” for different kinds of data storytelling
• Best practices for human-in-the-loop chart debugging
• A checklist for deploying LLM-generated visualizations
• Inspiration for turning prompts into dynamic data art

If you’re a data storyteller, designer, or a developer, this session will spark ideas on how to discover deeper insights from data and automate your data storytelling process using LLMs via real-life techniques.

Anand is an LLM psychologist at Straive. (It’s not an official title. He just calls himself that.) He co-founded Gramener, a data science company that narrates visual data stories, which Straive acquired. He has hand-transcribed every Calvin & Hobbes strip ever and dreams of watching every film on the IMDb Top 250.