iFeeling Daily
Daily curated AI insights you can't miss.
Baseten joins Hugging Face Inference Providers for serverless model access
Baseten joins Hugging Face as an Inference Provider, giving developers serverless access to models like DeepSeek V4 Flash via the Hub or SDKs with optional direct billing.
Mirendil signs $100M+ Google Cloud deal for self-improving AI
Mirendil has signed a $100M+ multi-year partnership with Google Cloud to scale its self-improving AI research, securing compute capacity with TPUs and GPUs.
Mirendil Selects Google Cloud AI Hypercomputer with TPU and NVIDIA
Mirendil, a frontier AI lab, selected Google Cloud's AI Hypercomputer, using a mix of TPU and NVIDIA infrastructure to accelerate AI research and development.
Spending Like a Hyperscaler
SpaceXAI spent $15.83B on AI infrastructure last quarter, funded only 12% by operations, while Microsoft covers 155% of capex from cash flow.
Klaviyo acquires Elias Torres’ Agency in full-circle reunion
Klaviyo acquires Agency, an AI customer success startup founded by Elias Torres, who will join as CPO. The deal brings together Torres and CEO Bialecki, who worked together at Performable in 2010.
UiPath built a shared GPU fleet on Google Cloud for agentic AI at scale
UiPath shifted from isolated GPU clusters to a shared fleet on Google Cloud, using A3 instances for training and G4 for inference to manage spiky workloads.
Solving the Noisy Neighbor with Sharded Architecture
By transitioning from a monolithic pipeline to a sharded hub-and-spoke architecture, platforms can isolate tenant failures and reduce blast radius from 100% to under 5%.
Filestore embraces Colossus for scalable AI storage
Google Cloud Filestore now runs on Colossus, decoupling capacity and IOPS while enabling high-concurrency NFS workspaces for AI agent swarms on GKE.
MacPaw Taps Liquid AI for On-Device Inference in SetApp App Store
MacPaw partners with Liquid AI to power locally hosted AI models, offering on-device inference to developers via SetApp app store.