Analysis
Watney, a San Francisco-based robotics startup building dual-arm robots for hyperscale data centers, raised an $80 million Series A co-led by Valor Atreides AI Fund and Hummingbird Ventures, with existing investors Conviction, Abstract and A*, plus Grant Gordon, also participating, according to WOWTALE.
The round brings Watney's total funding past $100 million, following a $21 million seed also led by Conviction, Abstract and A*.
Watney was founded in 2023 by Ryan Gannon and Sean Cheong. Its robots are built for cable replacement, power-cycling servers, reseating parts and inspecting equipment -- routine but labor-intensive maintenance work inside hyperscale facilities.
“The round brings Watney's total funding past $100 million, following a $21 million seed also led by Conviction, Abstract and A*.”
Inside Meta's Live Facilities
Meta is testing two dual-armed Watney robots on cabling work at its Altoona, Iowa campus, with related trials underway at its Prometheus data center in New Albany, Ohio, according to MLQ News. Meta's internal estimates suggest the technology could eventually replace up to 80% of certain technical maintenance tasks currently done by human technicians. Notably, Meta isn't testing Watney in isolation -- it's also running robots from ABB, the established Swiss-Swedish industrial automation giant, and Kinova, a robotic-arm maker, inside the same facilities, giving Meta a direct, real-world comparison across an incumbent, a specialist and a two-year-old startup.
The Competitive Field
Beyond ABB and Kinova, Watney's startup competitors include Swiss-based Exclaim Robotics, which raised $4.9 million in pre-seed funding, and Y Combinator-backed Shepherd Robotics, which is building a broader-purpose robotics foundation model rather than Watney's narrower, data-center-specific focus. Watney's bet is that specializing tightly on a repeatable set of maintenance tasks inside a single, well-understood environment beats a general-purpose robotics platform on reliability and speed to deployment.
The Numbers In Context
$80 million is a modest round set against this week's AI-infrastructure mega-deals:
- [Crusoe](/pulse/crusoe-3-9b-series-f-30-9b-valuation-2026) -- $3.9B Series F at a $30.9B valuation
- [CoreWeave](/pulse/coreweave-3b-convertible-bond-atm-program-2026) -- roughly $6B combined debt-and-equity raise
But data-center robotics sits in a distinct, physical-operations layer of the same buildout: someone still has to plug in cables and swap parts across the gigawatts of compute those larger rounds are financing.
What the funding headline misses: Meta's Iowa and Ohio deployments are pilots, not a signed commercial contract at scale, and the 80% figure is Meta's own internal estimate, not an externally verified benchmark. Cable-plugging and server-swapping inside a live hyperscale facility carry real failure costs -- a robot that damages equipment or triggers unplanned downtime could set adoption back further than a funding round timeline suggests, and Watney has not disclosed how its error rate compares with a human technician's over a comparable volume of tasks.
The open question is concentration: Meta is currently Watney's only disclosed hyperscaler customer across two facilities. Whether Microsoft, Google or Amazon show up as customers before Watney's next round will determine whether this is a genuine market or a single well-funded reference deployment.
What Founders And GPs Should Take From This
For founders building physical-AI or robotics companies, Watney's path -- a $21 million seed, roughly two years of narrow product focus, then a hyperscaler pilot before a $80 million Series A -- is a more capital-efficient route to validation than trying to raise a mega-round on a broader robotics platform pitch before any single customer has proven the unit economics.
For GPs, the read is similar to Comp AI's $34 million Series A this week: real revenue-adjacent traction at a modest check size is still getting funded even in a year when headline dollars concentrate in AI-infrastructure mega-rounds.