Profile of Start-up AirMettle

Company:
AirMettle, Inc.

Located in:
HQs Houston and other sites in San Francisco, CA, Washington, DC, Melbourne, Australia, and Tokyo, Japan

Web site:
AirMettle.com

Founded in:
2018

Financial funding:
Over $3 million in seed investments, and over $1 million in SBIR/STTR grants

Founders:
Donpaul Stephens: Founder and CEO, he is a serial entrepreneur best known as the founder of Violin Memory where he developed the original concept; hired and led the team; raised venture capital from strategic, institutional, and individual investors; established key strategic partnerships in both supply chain and go-to-market partners; and developed multiple accounts. He architected AirMettle’s approach for parallel in-storage analytics and has led the company’s development from inception to date.

Matt Youill: Co-founder and head of analytics, he was formerly chief technologist at Betfair, which he joined a team of 5 engineers, leading the re-architecture of firm’s systems to see it become one of the world’s biggest online gaming companies. He designed a ground up replacement for Oracle that would become one of the fastest databases in the world, processing more transactions than all the European stock exchanges combined and becoming the catalyst for the launch of Betfair’s own financial exchange LMAX.

Chia-Lin Wu: Co-founder and head of object storage; she has over 20 years experience developing large scale distributed systems. She led the development of a distributed object store at Pi-Coral. Previously, she was a senior engineer at AOL, leading development of dynamic reporting APIs and parental control back-end systems and content filters.

Employees numbers:
~ 20

Revenue:
Not disclosed at this time

Technology:
It integrates high-scalable analytics into software-defined object storage that works on standard hardware on-premise and in leading clouds. Highly-parallel processing of data in the storage layer accelerates analytics by up to 100x, leveraging underutilized resources in storage controller servers. Firm’s “pre-processes” (select, aggregate, re-scale, etc.) data in storage to return significantly smaller data objects to the analytics tier – reducing loads on and costs of networking, storage, memory, and compute for analytics.

Click to enlarge

Products:
AirMettle Analytical Storage Platform

Click to enlarge

Release and roadmap:

Product is in initial deployment with multiple early access trials underway (traditional event/record data analysis)
Multi-dimensional data (NetCDF4) features will enter trials in Summer 2024
AI inference capabilities will be demonstrated in late 2024 for trials in 2025

Pricing model and price:
Not disclosed at this time

GTM:
Direct and partners to enterprise
Government partners to Federal and State

Customers:
Los Alamos National Laboratory

Workloads/Use cases/Applications:

“Near Real-time” analysis of event data (logs, configurations, etc.)
Ad hoc queries on demand, no indexing required, without penalty
Multi-dimensional data analytics/reduction
Extract subsets of large climate models, rescale on demand to speed weather forcasting

Target market:

Enterprise: security, IT/network operations management
Manufacturing: semiconductors, energy
Scientific: materials, weather services (NOAA and “clients”),

Competition:
AirMettle was designed from the ground up to integrate and optimize analytics in the object store (a.k.a. data lake). Object Storage solutions traditionally were designed to optimize storage of semi-structured data, but not perform analytics; various data lake vendors have added basic analytics, but functionality and performance is limited by architectures. Data warehouse solutions were traditionally designed to accelerate repeated queries of well understood data stored in structured, often proprietary formats; various vendors have added some support for semi-structured data object formats, but again, functionality and performance is limited by architectures. By comparison, AirMettle has richer querying functionality (S3-compatible and more, with advanced APIs to better support analytical platforms such as Apache Spark) enabling the storage to significantly accelerate ad hoc analytics while reducing overall solution costs.

Click to enlarge