Alex Palcuie has joined the AI Reliability Engineering (AIRE) team at generative AI business Anthropic.
Palcuie spent the past eight years at Google, most recently provisioning new GPU capacity for the company's cloud service.
"I've joined Anthropic as a Member of Technical Staff on the newly formed AI Reliability Engineering (AIRE) team," Palcuie said on LinkedIn.
"So far, the vibes have been immaculate. I'll continue to be based in London, with Claude's reliability and outages gradually becoming my problem now. I’ve been impressed by the work the team has done so far, and by how seriously Anthropic takes reliability."
The AIRE team is led by Todd Underwood, who joined last November from OpenAI and Google.
In his own LinkedIn post earlier this year, Underwood said the team was "looking for staff+ (L6+ in Google/FB level scheme) people right now, especially those with large-scale distributed systems and even better large scale AI/ML experience. We'll broaden the scope of requirements soon but this is where we're starting."
The team is based out of London, UK, and San Francisco, US. A job listing for a San Fran-based AIRE staff software engineer has a salary band of $320,000-$485,000.
Anthropic works across multiple chip architectures, alongside the classic Nvidia-based GPU architecture used by most AI companies.
With Google owning around 14 percent of Anthropic, it also uses the firm's TPUs. Amazon Web Services has pumped around $8 billion into Anthropic, with the generative AI business using Trainium chips.
Amazon plans to build Rainier, a cluster of Trainium2 UltraServers containing hundreds of thousands of Trainium chips interconnected with third-generation, low-latency petabit scale EFA networking, to be used exclusively by Anthropic.
