Investing in Archil

Storage is a critical bottleneck limiting AI and big data performance. While GPUs and compute infrastructure have accelerated dramatically, traditional storage architectures haven’t kept pace. Teams use object storage like S3 for large, inactive dataset storage in the cloud, yet today's AI applications require active processing of more data than ever before. The existing technologies for working with this data (such as EBS) don't meet the scale, performance, or cost requirements of these modern applications, resulting in teams building complex ETL pipelines to avoid slow retrieval speeds. As a result, GPUs sit idle, workflows stall, and AI-driven companies miss out on critical performance gains. 

What if we could fundamentally rethink storage architecture to deliver object-storage scalability at block-storage speeds? This would enable AI teams to fully utilize their GPUs, eliminate costly pipelines, and achieve new performance heights.

Enter Archil: the intelligent data platform that powers modern AI applications that need instant access to datasets of any scale, with the same low costs and simplicity of cloud storage.Archil eliminates these barriers by making enormous datasets in cloud storage as fast and seamless as local disk, unlocking a new era of scalable, high-performance computing.

Archil’s core innovation is a unified data layer that connects compute directly to infinite cloud storage, without requiring manual data movement or capacity planning. Unlike traditional solutions, Archil provides EBS-level speed at a fraction of the cost by leveraging intelligent caching and a high-speed, distributed access layer. It’s 30× faster than direct S3 access, delivering sub-millisecond latency while ensuring data remains synchronized in its original S3 format. This means AI training, analytics, and high-performance workloads can access vast datasets instantly, without changing a single line of code.

The market for cloud data storage is growing rapidly—right now it’s a $37B+ opportunity and is projected to grow to $234B+ by 2031. Companies are racing to process more data, but legacy storage architectures weren’t designed for the scale and speed that AI and cloud workloads demand. Archil removes storage-induced friction, allowing enterprises to build truly stateless, high-performance applications without complex infrastructure. By eliminating cold starts and making massive datasets instantly available, Archil accelerates AI development and large-scale analytics in ways previously impossible.

What sets Archil apart is its ability to combine the simplicity of cloud storage with the speed of local SSDs, eliminating the operational headaches associated with managing storage infrastructure. Archil’s fully managed, cloud-native approach offers a plug-and-play solution that scales effortlessly while reducing costs by 90%. Customers like Depot and Fly.io are already working with Archil to simplify their data management because it's the only shared data platform with the high performance required for their end-users.

The Archil team is uniquely equipped to reimagine cloud storage, led by CEO Hunter Leath, who brings a rare combination of technical depth and practical experience in scaling storage systems. Hunter was a founding engineer on Amazon’s Elastic File System (EFS) team, where he contributed directly to the low-level architecture, an experience possessed by few engineers. He later transitioned into a senior product management role for EFS, navigating both strategic product decisions and highly technical engineering challenges as the system scaled. 

At Netflix, Hunter built upon his AWS foundation to optimize Netflix’s cloud storage performance, handling the streamer’s extremely high-throughput, latency-sensitive workloads. Hunter brings exceptional founder-market fit to Archil, combining deep technical expertise in cloud storage infrastructure with a proven track record of scaling mission-critical systems at Amazon and Netflix. Under his leadership, Archil has developed an engineering culture centered around rigorous benchmarking, continuous performance optimization, and rapid, data-driven innovation, positioning the company to redefine how enterprises approach cloud storage for AI and high-performance computing workloads.

Archil is building the next foundational layer of cloud storage, transforming how data is accessed and managed at scale. Today, we are honored to announce that Felicis led Archil’s $6.7M seed round. We are joined by Y Combinator, Wayfinder Ventures, Peak XV, General Catalyst, Lombardstreet Ventures, Twenty Two Ventures, Theo Browne, Erik Bernhardsson (Modal), Ryan Worl (WarpStream), Wayne Duso (formerly AWS), Amit Gupta (formerly Benchling), among others. We are incredibly excited to support Hunter and the entire Archil team in redefining the future of cloud storage. 

Join the Archil community! You can follow Archil on Twitter and LinkedIn. They are actively hiring; check out their opportunities here!