Akave raises $6.65M for compute-agnostic, egress-free cloud storage – News

Akave Inc., a two-year-old cloud infrastructure startup, is formally entering the enterprise storage market today with the launch of an S3-compatible object storage service that decouples storage from compute infrastructure.

The firm also said it raised $6.65 million in seed funding.

Akave Cloud lets enterprises move data across public clouds, neoclouds and on-premises environments at a flat-rate list price of $14.99 per terabyte per month, with no egress fees or data transfer charges.

The company was founded in response to growing enterprise frustration with vendor lock-in and unpredictable pricing models among hyperscale cloud providers, said founder and Chief Executive Stefaan Vervaet.

“Moving data from Google Cloud, for example, to one of the newer neoclouds, or bringing it back on premises, is very expensive,” Vervaet said in an interview. By eliminating egress fees and separating storage from compute, Akave aims to give customers more negotiating leverage and architectural flexibility.

Typical use cases range from processing large consumer intelligence datasets for advertising analytics to ingesting astronomical observation data and managing AI training datasets.

Sovereign and immutable

Akave is also targeting organizations facing rising regulatory and data sovereignty requirements, particularly in Europe. Vervaet said compliance demands, such as the General Data Protection Regulation and the National Institute of Standards and Technology frameworks, are pushing enterprises and research institutions to seek stronger controls over data residency, encryption keys and audit trails.

“Immutable audit trails, and the ability to prove who accessed what, is becoming more important, and that’s built into our stack,” he said.

Under the hood, Akave converts stored objects into cryptographic hashes based on file content, called Merkle trees, and records them on an immutable blockchain ledger. The design creates tamper-evident audit trails while keeping the actual data off-chain. The architecture is intended to support zero-trust environments and guard against data corruption or poisoning in AI workflows.

Akave breaks its platform into three components: storage nodes, the immutable ledger for cryptographic hashes and a gateway layer that translates standard S3 application programming interface calls into the underlying cryptographic structure. Customers can consume the offering as a cloud service or deploy a self-hosted gateway to maintain control of encryption keys and metadata placement.

Rather than building and operating massive proprietary data centers, Akave relies on a distributed network of infrastructure providers, many of which emerged from the blockchain ecosystem, that supply storage, networking and compute capacity in localized markets across the U.S., Canada, Europe and parts of Asia. “We decouple infrastructure from services,” Vervaet said.

Snowflake qualification

Akave said it has been qualified as an external storage provider for Snowflake Inc. and is integrated with the Apache Iceberg table format. It said one marketing technology company migrated from Amazon Web Services Inc.’s Athena, and S3 to Snowflake for analytics, while using Akave as the underlying storage layer.

By storing data on Akave and running computing separately, the customer can shift workloads among AWS, Google Cloud or on-premises systems without paying network transfer penalties, Vervaet said.

The platform supports object lock, versioning, identity and access management policies and bring-your-own-key integrations. It also offers programmable, blockchain-based access controls and on-chain verifiability for auditability.

AI-focused storage layer

Akave is not positioning itself as a provider of computing services or an artificial intelligence model training platform. “We’re compute-agnostic,” Vervaet said. “We’re focusing on one particular part of the stack, and that’s storage.”

He said data transfer costs are often the second-costliest part scaling AI workloads, particularly when organizations attempt to move large training datasets between cloud environments.

Investors include CVP NLH Advisors Ltd, which does business as No Limit Holdings, Protocol Labs Inc., Big Brain Holdings LLC, Avalanche VC LP and the Filecoin Foundation. The funding will be used primarily to expand sales and marketing, with roughly 80% of the company’s 15 employees currently focused on engineering.

Image: News/Microsoft Designer