Sequoia Capital just bet $45 million on an idea that sounded like science fiction until yesterday: an artificial intelligence that runs live product demos without a human. Sable, a startup less than a year old, unveiled Aidan, a system that not only showcases software features but answers potential customers’ questions in real time and switches languages mid‑conversation. The ambition is clear: replace not just chat support, but the entire role of the demo presenter.

The funding, led by Sequoia and 8VC, marks a symbolic shift. We’re no longer talking about chatbots answering predefined FAQs; these are autonomous agents taking control of a complex sales flow. Yet the very nature of these demos raises a question that the round numbers of the investment can’t mask: who safeguards the data that the AI processes during a demonstration? Every product demo touches potentially confidential information — future roadmaps, technical details under embargo, prospect data. In many regulated industries, merely processing that information on a public cloud can breach corporate policies or regulations like the GDPR.

This is where Sable’s trajectory intersects the themes close to AI‑RADAR’s heart. Automating demos pushes companies to rethink AI deployment not only in cost terms but from a data sovereignty perspective. If an agent like Aidan had to run on proprietary stacks to prevent leaks, precise hardware constraints would emerge: acceptable latency for real‑time dialogue, sufficient inference capacity without noticeable buffering, all in a self‑hosted setup. The difference lies in the technical specs that no venture capital press release ever mentions: model quantization, VRAM requirements, throughput in tokens per second.

For those evaluating such deployments, familiar trade-offs come into play: a cloud service slashes initial CapEx but ties data to external providers; on‑premise infrastructure guarantees control but demands GPU investments and ongoing maintenance. The choice is far from trivial.

Today’s announcement isn’t just a capital injection. It signals that the market for autonomous demos is about to explode, along with the tension between those promising convenient cloud AI and those needing to keep it in‑house. If Sequoia is betting so heavily on Sable, it’s because they see a future where every company will want an “AI employee” for sales. But that future will only scale reliably when on‑premise infrastructure can bear the load without forcing compromises on privacy.