We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!
AI adoption levels are higher than they’ve ever been in the enterprise. According to a January survey conducted by Gartner, corporate use of AI grew 270% over the past four years. But developing, deploying, and managing AI applications at scale requires a platform that supports doing so, which is what startup company Iguazio provides. Its investors believe it has legs: Iguazio announced that it has secured $24 million in a funding round led by INcapital Ventures with participation from existing and new investors including Samsung SDS, Kensington Capital Partners, Plaza Ventures, and Silverton Capital Ventures, bringing its total raised to around $75 million following a $33 million series B in July 2017.
“This is a pivotal time for AI. Our platform helps data scientists push the limits of their real-time AI applications and see their impact in real business environments,” said CEO Asaf Somekh, who cofounded Iguazio in 2014 with former senior software developer and architect Orit Nissan-Messing and Voltaire colleagues Yaron Haviv and Yaron Segev. “With support from INcapital, Kensington Capital Partners, and our other investors, we are ready to expand our international team and reach our ambitious goals.”
Iguazio’s product suite collects data and preps it offline or offline, accelerating and automating AI model training for deployment via APIs. Event-driven streaming, time series, NoSQL, SQL, and files are among the types supported, which can be explored and manipulated by a real-time data layer that lets admins use a variety of protocols and concurrently read the data with third-party analytics and data science frameworks.
Nuclio, Iguazio’s open source serverless framework, streamlines machine learning pipeline steps like packaging, scaling, tuning, instrumentation, and continuous delivery with features like rolling upgrades, A/B testing, logging, and monitoring. Its MLRun open source framework parallelizes work within a single pod by wrapping engines around tools like Spark, Google’s Tensorflow, Horovod, and Nuclio, and it handles a range of triggers including HTTP and cron.
The Iguazio suite — which runs as in-memory databases on flash memory — continuously trains models in a production-like multi-cloud or hybrid cloud environment, dynamically scaling graphics cards, processors, and memory and automatically tracking code, metadata, inputs, and outputs of executions in a reproducible fashion. (Admins can track the elements of all running jobs as well as historical jobs and store them in a single report.) The platform can run multiple experiments simultaneously and select the best model, and it facilitates the migration of this model from an integrated development environment to production.
Enterprise customers get an integrated offering they set up through wizards that help configure administration policies and notifications. Managers securely share data by providing direct access to it, so that employees are exposed to different elements according to predefined rules. The data layer classifies transactions with a built-in firewall that provides fine-grained policies to control access, service levels, multi-tenancy, and data life cycles, and that enables organizations to collaborate and govern across apps and business units without compromising security or performance.
There’s no shortage of orchestration platforms in the over $1.5 billion global machine learning market. Amazon recently rolled out SageMaker Studio, an extension of its SageMaker platform that automatically collects all code and project folders for machine learning in one place. Google offers its own solution in Cloud AutoML, which supports tasks like classification, sentiment analysis, and entity extraction as well as a range of file formats, including native and scanned PDFs. Not to be outdone, Microsoft recently introduced enhancements to Azure Machine Learning, its service that enables users to architect predictive models, classifiers, and recommender systems for cloud-hosted and on-premises apps, and IBM has a comparable product in Watson Studio AutoAI.
That’s not to mention rival startups like San Francisco-based Databricks and Explorium, which provide suites of enterprise-focused scalable data science and data engineering tools. But Somekh is confident in the robustness of the Iguazio platform, which he claims prevents 60% of faults and serves up to 500 real-time recommendations per second.
To this end, one new customer — digital payment platform Payoneer — recently deployed a proactive fraud prevention solution across four million accounts using Iguazio’s tools.
“Iguazio’s unique technology facilitates the data science creation process from start to finish, enabling enterprises to deploy AI applications that create real business impact,” said Kensington Capital Partners chairman Tom Kennedy. “With the opportunities we are seeing for machine learning technology and the global success stories emerging from this high-tech nation, Iguazio represents a great first investment for Kensington in an Israeli company.”
Iguazio is based in Herzliya, Israel, and has offices in the U.S., U.K., and Singapore.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn more about membership.