Sable announced $45M on July 16, 2026, co-led by Sequoia (Shaun Maguire) and 8VC (Joe Lonsdale, both joining the board), for “Aidan” – an AI employee that runs live product demos, qualification, and onboarding by actually seeing and clicking the product in a shared virtual workspace while talking to the buyer. Less than a year after founding, the company says Aidan is in production at Notion, Decagon, and unnamed large public companies. Note the sourcing: customer names, capabilities, and traction are all vendor-stated; the round has no stage label and no disclosed valuation.
Our read: this is the most credible attempt yet at the AI sales engineer, a different – and we think more defensible – job than the AI SDR. Our churn data showed packaged AI SDRs failing because outbound trust-building is judgment work; a demo, by contrast, is a bounded, product-grounded interaction where infinite patience and 24/7 availability are genuine advantages and hallucination risk is constrained by the product on screen. The bet has two sharp edges, though. Technically, low-latency live computer use is frontier capability – a frozen screen mid-demo in front of an enterprise buyer is a category-defining failure mode. Strategically, Sable sits directly in the blast radius of foundation-model vendors shipping computer-use agents natively. The Sequoia/8VC board commitment and the Decagon logo (an AI-native vendor choosing to buy rather than build this) are the strongest external signals we have. Watch for named customer references and published pricing before treating this as a category, not a company.
Strengths
Genuinely new interaction pattern: real-time computer use + vision + voice in a shared LiveBox VM, so the agent clicks and navigates the actual product while talking - not a chatbot over docs; production customers include Notion and Decagon (company-claimed); Sequoia (Shaun Maguire) and 8VC (Joe Lonsdale) co-led and took board seats; strong research-heavy founding team out of Harvard with SpaceX/Google/Meta/Together AI backgrounds.
Weaknesses
Less than a year old with no disclosed pricing, revenue, or valuation; the round itself carries no stage label; low-latency live computer use is a frontier capability - failure modes in front of a live prospect are the highest-stakes kind; every capability claim is currently vendor-sourced.
Opportunities
The AI SE / demo-automation slot is far less crowded than AI SDR; presales headcount is expensive and demo capacity gates pipeline at most B2B companies; improving vision/computer-use models raise its ceiling every quarter.
Threats
Foundation model vendors are shipping computer-use agents natively and could commoditize the core capability; incumbent demo-automation tools (Consensus, Reprise-class) and conversation platforms can move up; a public failure during a live enterprise demo could set the category back.