Hex, an agentic data platform, is using GPT‑6 Astra to turn automated analyses into interactive, shareable visualizations. “A really important part of data work is not just doing the analysis, but also communicating it,” says Caitlin Colgrove, co‑founder and CTO of Hex. She argues that models have long struggled with data visualization, and that GPT‑6 Astra represents an advance in this area.
What GPT‑6 Astra brings
Creating a useful visualization requires both technical implementation and presentation choices. Colgrove notes that GPT‑6 Astra works well with the underlying libraries and performs the necessary data transformations, helping analysts produce functional and visually appealing outputs.
For example, a Hex user can ask how sales channels are performing over time, which channels are doing best, and which might be discontinued. Hex delivers written findings alongside trend lines and comparisons, and provides interactive dashboards that let users explore key metrics, rankings, and geographic breakdowns.
Verification and relevance checks
Even technically correct code can yield answers that miss the point, so Hex uses GPT‑6 Astra to interrogate results: Do the numbers make sense? Does the analysis answer the user’s question? Does it reflect the business objective?
Colgrove says GPT‑6 Astra investigates more thoroughly, uncovers more of the data’s pitfalls and nuances, and produces a more comprehensive analysis than earlier models.
Practical impact
A clear visualization gives colleagues a way to engage with analysis—comparing results, tracking changes over time, and exploring the details relevant to their work. With GPT‑6 Astra, Hex users can more easily access, understand, and communicate data, advancing Hex’s mission of making everyone a data person.
