Executives Run AI Models Locally on Private Data

A Shopify CEO's wine-cellar setup shows how individuals can keep sensitive information off outside services.

Updated on Nov. 2, 2026 in Artificial Intelligence

Executives Run AI Models Locally on Private Data

Live Poll

Would you be comfortable with your employer tracking your heart rate and calendar data?

Shopify CEO Tobi Lütke has converted his wine cellar into a computing room that runs AI models locally, analyzing personal data from a Fitbit and his calendar. The setup is a personal deployment rather than a commercial product; it is not described as available to others.

Why it matters

Running models locally keeps sensitive information off external services, a trade-off that matters to executives handling confidential material. It also shows how individual leaders are testing AI for personal productivity as companies push staff toward AI use.

The setup pairs a Dell GB300 system built on Nvidia's Grace Blackwell Ultra platform with a script that pulls Fitbit heart-rate, calendar and location data. The source provides no throughput, power or cost figures to compare it against other hardware.

The players

Shopify

Canadian e-commerce platform company whose CEO Tobi Lütke is pushing staff toward AI adoption.

Tobi Lütke

Shopify CEO who built a local AI computing room in his wine cellar to analyze his own calendar and health data.

Nvidia

Chipmaker whose Grace Blackwell Ultra computing platform underlies the Dell GB300 system used in the setup.

Dell

Hardware maker that supplies the GB300 system running the AI models in the wine-cellar room.

The details

The room holds computing hardware that runs AI models locally, meaning the models execute on the owner's own machines rather than on a remote service. A script gathers data from a Fitbit wearable and his calendar, and the system compares heart-rate readings against meeting schedules to flag which meetings coincided with the most stress. Lütke said he ran the system locally because he handles sensitive information he cannot send to outside providers.

Timeline

  1. In April 2025, an internal Shopify memo from Lütke urging greater AI use was made public.

The Tech Race

The Grace Blackwell Ultra platform is one of the high-end systems Nvidia offers for AI workloads, and this setup shows it being used in a single personal room rather than only in large facilities. The source does not place the deployment against any competing hardware or research program.

Individuals with sensitive personal or professional data can run AI models on their own hardware, avoiding sending that information to outside services, though the source gives no price or setup timeline. The source describes this as one executive's personal deployment and does not say it is offered to others.

The takeaway

Local AI setups like Lütke's show how individuals can keep private data off external services by running models on their own hardware. Readers can watch whether companies like Shopify publish guidance on local AI use for staff, since the source gives no further milestones.

Further reading

For broader context on how leaders are adopting AI tools, explore our Artificial Intelligence coverage.

Source note: This article includes information reported by Hindustan Times.

Live Poll

Would you be comfortable with your employer tracking your heart rate and calendar data?