Exclusive Interview With Asaf Somekh, CEO and Co-Founder, iguazioData stack design to accelerate performance in big data, IoT and cloud-native apps
By Philippe Nicolas on 2017.02.28
Who is Asaf Somekh?
- CEO and co-founder, iguazio
- VP marketing and business development, Compass-EOS
- VP, solution marketing, Mellanox
- Various executive positions at Voltaire
- Director R&D, Nexus Telocation,
- Software engineer/manager at Aptel
StorageNewsletter: Could you summarize the genesis and background of the company? Year founded? Number of employees?
Asaf Somekh: iguazio was founded in 2014 with a fresh approach to the data challenges faced by today's enterprises. Building a data pipeline has become so complex that the market now requires change and we’re leading the charge. We’ve assembled an experienced team to fundamentally redesign the entire data stack. iguazio bridges the enterprise skill gap and accelerates performance in big data, the IoT and cloud-native applications. We have more than 50 employees in the USA and Israel and are currently focused on our go-to-market strategy, actively hiring to expand across the board.
We know Yaron Haviv and you. Who are the other people behind the company? What are the products, big data or storage technologies they already developed?
We're fortunate in having a diverse leadership team of seasoned innovators in big data, cloud, storage, machine learning, high frequency trading, high performance computing, analytics and networking. Yaron Segev is a third founder, in addition to myself and Yaron Haviv. Segev is a serial entrepreneur who founded XtremIO, which was acquired by EMC in 2012. The three of us previously worked together on high-performance computing and datacenter solutions at Voltaire and Mellanox. Other members of our team came from EMC, IBM XIV, Cisco and Avaya. It's a great mix of people who have built big data platforms at companies like AT&T and Orbitz.
How is the company capitalized?
iguazio is backed by top VCs and additional strategic investors.
What's the company vision?
While many companies 'solve' the complexity of data by hiding the problems, iguazio’s vision is redesigning the data stack from the ground up with our high-performance unified data model architecture. Our mission is to enable the digital transformation by helping our customers share, consume and analyze data faster and with greater simplicity. Our resulting integrated solution is as simple and agile as the cloud, but its efficiency is magnitudes better for addressing the big data challenges of today and the future.
What are the challenges you wish to solve?
We're addressing the need for simplicity, performance, security and agility. Most enterprises today can’t truly leverage the IoT because they struggle with the adoption of big data, suffering from too many data silos and long transformations between data models. I want to be specific about security, because when we talk about it as a challenge what we really mean is data sharing. Today, most companies can't restrict access to data using user credentials, so they copy it again and again and again for different data models and platforms. That makes the data less secure, more expensive and hugely inefficient. iguazio’s unified data model provides simultaneous access through multiple industry-standard APIs: stream, object and file data are all stored in a normalized form only once, eliminating the need for copies.
What are the target markets?
We’re working with enterprises in the USA, Europe and Asia. Our beta customers are in the USA, London and Singapore in industries such as financial services, IoT, connected vehicles and cloud-based data services.
What kind of product is it? We heard it’s a mixed of several approaches?
You heard correctly. iguazio has taken a revolutionary approach; our Enterprise Data Platform is an entirely new means for addressing the data stack. It leverages innovation in storage, next-gen databases and security, and includes strong characteristics related to self-service and manageability that are often associated with cloud solutions.
What is the Unified Data approach?
Our unified data model lets customers store and index all kinds of data once and then access or manipulate it simultaneously using different models and apps for mobile, dashboards, analytics, machine learning and more. This means that users can perform real-time insights and actions, while significantly cutting deployment time and complexity.
How do you leverage open source?
We are maintainers, contributors and users of open source - it's at the core of our strategy. Because our approach is brand new, we’ve had to develop generic technologies for messaging, storage, security and benchmarking. We’ve already made some of them public in Github and will add more later this year (such as Accelio RPC, an HTTP Blaster and our NVMe based K/V store). We believe open source still leads to lock-ins and a greater focus on standardization is a must, so we will continue driving standardization of modern data APIs like our v3io API library. In the future, we will open v3io as well as the plug-ins we've developed for Spark, Hadoop, NGINX, Kubernetes and others.
In fact, the company develops a product based on solid technologies. Could you elaborate on a few of them?
The core of our system is an innovative, high-performance unified data model. That allowed us to redesign the data pipeline, providing a way for users to efficiently consume data simultaneously through a variety of APIs on top of a single repository. iguazio leverages 100GbE RDMA technologies, CPU parallelism, next-gen NVMe and NVM to secure, store and manage data and metadata at extreme speeds. We partner and work closely with vendors such as Intel, HGST and Samsung that deliver NVMe and non-volatile memory solutions.
What is your plan for other data access protocols and interfaces?
Our system is built so it is easy to add more APIs. At the same time, we continuously enhance our analytics engine processing capabilities to offload higher-level data processing tasks. We plan on releasing our 'server-less' processing framework which will allow customers to rapidly build their own APIs and event-processing tasks. We will also expand our support to more database frameworks like Apache Presto and other ANSI SQL tools, and will add IoT focused time-series and messaging APIs. We hope to see partners and customers adding and contributing new interfaces once v3io APIs are open.
What about new Apache projects? Are you able to include and integrate them pretty fast? What about expertise around these?
We've got several Apache, Spark and Hadoop experts on our team and iguazio is fully integrated with Spark, Hadoop and Kafka clients. We'll also offer Presto support soon. Integration is easily completed by contractors within the community. We are working on v3io Java, Scala, Go and Python libraries to make it simpler to use and further optimize performance.
How do you manage back-end object storage? File services?
Legacy databases are typically block-based and they rely on locking, which limits performance and scalability. Our approach is different. We map all data as key/value store and have key/value implementations for native NVMe, file or object storage. The database engine, along with our data lifecycle management tasks, controls tiering and data movement between various storage pools.
What about competition?
Most other vendors mask problems through various caching solutions. That solves nothing. In contrast, iguazio enables cloud providers to compete with Amazon and provides enterprises with an alternative. We've partnered with cloud and managed services providers. With iguazio, rather than being locked in to Amazon APIs, enterprises can use these APIs in hybrid and on-prem environments. Our APIs are compatible with Amazon and improve it, so the result is much better performance without Amazon's lock-in. We don' compete with traditional storage or database companies. Rather, we extend an AWS-like platform for the enterprise market and for edge applications.
What makes the company unique?
iguazio is unique because we approach data differently. We’ve created a new type of data platform that provides a completely fresh approach to handling data. Rather than hide or work-around the problems, we address and solve them with our unified data model.
We've also got the best minds in the business working with us. Together, we've designed a unique architecture and built our solution from the ground-up to maximize CPU utilization, leveraging non-volatile memory, flash and dense storage to achieve extreme performance at the lowest cost.
Your business model?
We're already working with channel partners and system integrators to target enterprises - both established enterprises undergoing the digital transformation and new players that are disrupting industries.
How do you sell?
We're working closely with key end users in a variety of sectors: financial, healthcare, connected vehicles, ride sharing and IoT. We also recently announced our partnership with Equinix for hybrid cloud solutions. We actively partner with Intel, Dell and others for both technology and sales purposes.
How many customers?
Our customers are diverse and represent large financial exchanges, banks, cloud providers and a leading ride-sharing company. While contractual obligations currently prevent us from naming them publicly, we'll provide more information about them in the coming months, including detailed case studies about the impact of iguazio’s platform on their bottom lines.
How is the product priced and licensed?
The iguazio Enterprise Data Cloud platform is available either through software licensing or as an appliance. We offer an annual subscription license per server. Appliances are priced separately as a one-time fee.
What about the international presence?
We're already global, with customers in the USA, Europe and Asia. Our American team is expanding quickly and we're expecting growth in Singapore and London as well.
What are the priorities for 2017? What is the future for iguazio?
Our focus this year is on enterprises which require agility and the ability to adopt new technologies like cloud-native, real-time analytics, containers and serverless functions. We anticipate announcing the platform’s general availability over the summer and providing a support infrastructure to go with it.
Israeli Stealthy Start-Up Iguazio Raises $15 Million in Series A
In data management and storage solutions for big data, IoT and cloud applications software
2015.12.11 | Press Release | [with our comments]