Exclusive Interview with Valery Guilleaume, CEO, Nodeum

Valery Guilleaume, 44, is CEO of MT-C since January 2014 and Nodeum since 2016. He also is chairman of the board of MT-C since 2012, year of the acquisition, and spent time before that at various financial companies working on M&As, Taxes and at TI Automotive in various IT, security and audit positions as well. He’s the father of 3 children and is passionate about boat sports, sailing, rowing, and yachting, owning a yachting license. He also like do-it-yourself construction and building wood furniture.

StorageNewsletter: Nodeum is a Belgium company, founded in 2016, but a bit confidential, could you give us a company snapshot with some financial figures, employees, offices, indication of revenue range… with founders’ background as well?

Valery Guilleaume: Nodeum is engineered by MT-C, a privately-held enterprise software company headquartered in Belgium with its main offices in Liège.

My desire for entrepreneurship started when I was overseeing WW IT operations for an international automotive company that had a complex infrastructure spread across 130 sites, including 3 main data centers.

As most of the production operations were located at the edge, the need for data mobility and archiving was crucial. I realized the limitations of the current storage technologies to handle this mobility including the archiving, and protection of the data.

This led to the decision to acquire a company named MT-C in 2012, with my wife Jessica, who has been handling the financial and marketing aspects of the business since the beginning. This company already owned a software, LMS, well known in the IBM I and mainframe market with partnership with major vendors.

This acquisition of MT-C allowed us to bootstrap the development of new concepts which will be the genesis of Nodeum. It provided us with the necessary resources and credibility to transform the idea of Nodeum into a fully functional product. We bring to life Nodeum, a product that helps companies become more independent and flexible in solving their business challenges.

As one of the main co-founders of Nodeum, my primary focus is on the product aspect of the company. This involves overseeing the entire product development process, from conception to product delivery. To ensure that Nodeum meets the evolving needs of the market, we adopted an agile development approach. This allows our team to be flexible and responsive to market demands, while also ensuring that each feature is validated with key customers from the market before it is implemented.

During the development phase of Nodeum, we worked tirelessly to ensure that the software was built to the highest standards of quality and reliability. Every feature of Nodeum is meticulously crafted, tested, and validated with customers to ensure that it met their needs and expectations.

Throughout the development process, Jessica and I worked closely together to ensure that the financial and marketing aspects of the company were in sync with the product development. She played a vital role in managing the finances of the company and ensuring that the budget was allocated effectively to support the product development.

Today, Nodeum is a thriving enterprise software company that is trusted by businesses around the world to help them manage their data effectively. The success of Nodeum is a testament to the hard work and dedication of the creator and Jessica, who continue to drive the company forward with their complementary skills and expertise. We have 7+ employees and external consultants and we have generated less than $10 million last year, doubling each year for a few years.

How did the project start? What was the pain you wished to address? And what does Nodeum mean?
Nodeum is the concatenation of node and um. Node refers to the term “node” and “um” is the Latin neutral ending. The reference to the term “node” reminds us of the idea around the compute node, the storage node, and therefore distributed architecture systems.

The story of Nodeum started from a personal experience.

In 2014, we found out that our one-year-old son has a rare disease “Lymphangioma”. In the first stages of this discovery we realize that hospitals were unable to share patient’s data (MRI,…) to other hospitals without the support of the IT department. At the end, the only way which was adopted was to circulate with a copy of the images on a CD-ROM.

From there, we understand that the data cannot be easily transmitted across hospitals, which we find absurd.

Based on this statement, we extended our discussions with research centers to understand that specific systems such as DNA sequencers generate a lot of unstructured data and that no product on the market exists to quickly offload these generated contents to store and archive them efficiently.

To maximize their productivity, these systems must be used to the maximum of their resources. It is therefore necessary to be able to automate workflows and pipeline in moving back and force the data between where the data must be computed and where the data are stored.

We created Nodeum to help researchers be more efficient at putting their data where they need it. Their job is to keep their focus on their day to day of researcher, professor, etc.

We ambition to provide software to help them to increase their efficiency and automate the movement of data across all different storage.
We want to simplify and facilitate the day to day of these research centers by allowing them to connect directly and transparently to their data directly generated by sequencers, microscopes and others … on the powerful computer allowing them to analyze ….

We quickly realized that the problem of positioning the data could be encountered by other industries such as post-production which must archive very large volumes of data (400TB per film). The quality of the data does not allow them to compress the data. The delivery of the film master must be archived on several media (tape, cloud)

This is also easily extended to earth observation, CCTV, universities, ….

And by now, we see that supercomputing centers use data movers like Nodeum.

Our customers trust us because we can understand their needs and help them to maximize the usage of Nodeum to integrate into their workflows; and not ask them to change their workflow to cope with our software.

Nodeum is now based on modern technologies such as microservices, REST APIs, and plugins. Thanks to our dedication and hard work, we have successfully positioned ourselves and Nodeum in significant data management projects where major vendors like Dell, Quantum, Oracle, and HPE are keen to collaborate with us.

We want to be different from these traditional offerings by providing a complete key solution across hybrid storage and avoid to all customers a vendor lock-in scenario.

How is Nodeum financed? Did you get any VC money or is it just founders’ dollars? If not, do you plan to?
We didn’t get any VC money, yet.

Initially, Nodeum’s development was bootstrapped, meaning that the company was self-funded without the assistance of any VC investments. The co-founders, sales revenue, and government subsidies have all contributed to the financial support of the company during its early stages. Additionally, Nodeum has successfully raised funding from rounds of friends and angel investors where co founders participated too.

We are still available for VC to scale and expand Nodeum into the market and specially in the North America region.

Nodeum plays in the vast data management segment, how do you define what you do and what is the mission of the company?
Our primary focus is to simplify the process of archiving, protecting, and moving large volumes of unstructured data. We understand that organizations often face significant challenges when it comes to managing their data, especially when it comes to dealing with large volumes of information that are dispersed across multiple storage systems. That’s why we offer a unified platform that allows organizations to manage their data more efficiently and securely.

To achieve this goal, we work closely with our clients to integrate their data management plan into Nodeum, ensuring that they always have complete control over their data. Our platform is designed to accommodate large volumes of data, making it ideal for organizations that need to manage terabytes or even multi petabytes of information.

One of our key areas of focus is the end-user workflow and pipeline. We understand that managing data can be complex, which is why we strive to make the process as streamlined and user-friendly as possible. Our platform is designed with the end-user in mind, so that individuals can easily interact with data mobility and move data between storage systems with minimal hassle.

Specialists recognize Nodeum for its open data mover but you do much more than that. Let’s dig a bit into your product, I understand you just introduced a new major release. Could you detail this new offering?
We make a total redesign of the architecture in keeping the same experience that is appreciated from the user and the market.

The Nodeum is based on Linux OS, in this version we redesign its architecture to be micro-service oriented, with the capability to be extensible in using plug-ins to facilitate data access through storage-oriented connectors. This architecture provides end-to-end parallelization of each service to scale the solution horizontally and vertically.

As a result, the solution can scale its throughput and its metadata management capability to address hundreds of terabytes per hour of data transfer and to handle billions of files management in terms of metadata. Objectif is to be ready for exascale computing throughput requirements, sustain range of thousands of requests per second, handle millions simultaneous transfer requests, and throughput of hundreds of terabytes per hour.

We also aim as a 3rd objective to extend customer integration capabilities, the solution now allows event-based hooks registering, this allows each user to perform their own code and script execution in connection with the data movement operations.

In addition to the already available interface such as the Console, which is a GUI HTML5, we answer the demand to provide bash clients which are designed on the top of the same REST API services as the Console. This bash client, we named, ND Client is a cross platform client which allows any authorized user to handle data movement operations across all available storage.

And finally, this version supports its integration with the leader workload manager in Supercomputing which is SLURM. This is the most important integration for such a data management service in a HPC context.

Do you use any open source module or product?
Nodeum uses some open-source components.

• Kafka for message queuing processing across all nodes of the cluster
• MongoDB for Scalable Metadata
• Etcd and Zookeeper for Configuration

What are the use cases you target?
Our main target is a very high level of throughput and capacity management. We can find this level of requirement in HPC/supercomputer, this can include integration requirements with Workload manager such as SLRUM, Torque, PBC, etc.

Any references you can share? Do you know which product your users compare Nodeum with?
The most important HPC/supercomputing – research centers in Europe deployed Nodeum as Central Data Mover such as JSC, Germany, CSCS, Switzerland, BSC, Spain, CINECA, Italy and CEA, France.

What about your cloud strategy?
The strategy of Nodeum for the cloud space is:

1. Provide data mobility to easily migrate content to cloud instances with the objective of performing calculations and data processing…. Once the work is done on the instance, automatically retrieve the results to the main site or to perform an archiving.
   Because companies need to use the best of the resources they need. Traditional IT operations will become managed by Nodeum.
   In the cloud offering, we can also find services that will perform operations on the datasets, which means the data must go back and forth between on-premises services and cloud storage.
   In addition to cloud storage, we also recognize that there are many cloud services that can help organizations perform operations on their data sets. For example, many companies use cloud-based machine learning services to analyze their data and extract insights that can help make business decisions. However, this often requires data to be moved back and forth between on-premises services and cloud-based storage solutions.
   We recognize this challenge and work with our customers to ensure that their data is moved securely and efficiently, with minimal disruption to their operations.
2. need and requirement to use cloud-based storage targets to deliver content. This is especially true in the media industry, where companies need to be able to distribute their content quickly and easily to audiences around the world.
   By relyiTo deliver contents, especially for streaming players like Netflix or Amazon Prime for example.
   We are convinced that the cloud offers tremendous opportunities for companies of all sizes and more specifically in the field of media streaming.
   We believe our platform can help companies move their final content to AWS and other cloud storage solutions securely and easily.
   In working with our customers, we recognize that there is often a ng on cloud-based storage solutions, companies can take advantage of the scalability and flexibility offered by the cloud, making it easier for them to manage their content and meet the demands of their customers.

In conclusion, we believe that the cloud represents a huge opportunity for organizations to streamline their operations, reduce costs, and improve their ability to deliver high-quality services to their customers. Regardless of the industry, Nodeum can help you leverage the benefits of the cloud by providing a secure and reliable platform for the transfer and management of your data.

How do you sell the product? Any specific partners to mention?
For the USA and Canada, we started an agreement last month with Climb Channel Solution to accelerate the relationship with resellers like CDW Canada.

In Europe, we sell directly except in DACH where we have a sales agent to sell Nodeum, and in the UK where we use distribution.

In AsiaPac, we have also started to establish partnerships with key resellers in Japan and in India.

What do you think about S3 to tape market development? I understand Nodeum develops this kind of software services for some key players, any details on that?
Our firm is a highly experienced data management solution provider with a strong expertise in data archiving on tape. In fact, tape storage has been an integral part of Nodeum’s genesis, and we have continued to support it ever since. Our platform natively supports the Linear Tape File System (LTFS) protocol, making it easier for our customers to archive their data on tape.

Nodeum is certified by the LTO Consortium, and we support all major vendors such as Oracle, HPE, Quantum and more.

With the general adoption of S3 protocols, we make the step to also provide a S3 frontend in the solution to allow customers to ingest the data from a standard s3 gateway and offload in the backend data to tapes, this still in keeping the DNA of Nodeum. End-users can interact with the solution in using standard s3 commands from the push of the data to tapes to the data retrieval from tapes.

The difference with other vendors is that we always store in human readable format files on LTFS tapes. This to keep a no vendor lock in strategy.

At Nodeum, we understand the importance of the combination of the standard s3 protocol and the tape storage. We believe that tape storage can be a highly effective and secure option for long-term archiving. And as value add, we provide automation to directly transfer data from S3 to tape.

This solution is widely used by our key partners, including large universities, such as the University of Rostock, (10,000 students and employees). We help them to handle 3-2-1 rules where 2 copies of data are stored in S3 object storage and 1 copy is stored on tape storage using S3 as unique protocols.

What is the associated pricing model?

• Business. This is a subscription-based model: One price per year based on the capacity stored into the secondary storage. The customer pays effectively what is consumed. As a subscription model everything is included (capacity usage, support, onboarding, training).
• Enterprise: Enterprise is a perpetual license based. the license is based on the capacity stored into the secondary storage. Additional maintenance support fee as to be added on top of the license.
• Enterprise plus: This is equivalent to the enterprise model, but the license is based on the number of nodes.

Why should users prefer Nodeum vs. other solutions?
For these 3 main reasons:

1. Ability to simplify storage management and avoid vendor lock in.
   Nodeum provides a unified platform that simplifies the management of different storage systems including object storage, file-based storage, and tape library. Nodeum always uses standard protocols so that we never block a user with our solution and technology. Unlike other vendors, we use open writing formats to write content to any type of storage. In addition, the structure and content are always human readable.
2. REST API integration
   The Rest API integration capabilities facilitate fast integration and enable the automation of business operations chains. This process eliminates the need for internal IT department scripts, streamlining the transfer process, and avoiding IT pperation layer complications.
3. One single interface to manage and execute data mobility across all different storage.
   Reporting and task auditing are available to help the user to control data movement and detect mistakes on files.

What are your next steps? What can we expect for the rest of 2023?
We have started the development of the highly scalable virtual file system solution that will embed active archive workflows. We aim to address this user case in verticals such as banking who are in the transformation process of their business unit. The business unit will be the owner of the data and will be autonomous to manage their data, instead of relying on the IT operation.