INTRODUCTION

In 2021, we started XONAI with the mission to unlock the power of hardware to slash data infrastructure operational costs. Today, we’re delighted to announce that we have built an amazing team, achieved incredible breakthroughs in infrastructure optimization, and raised $3.5 million in seed funding led by Kadmos Capital.

Infrastructure optimization at the pace innovation demands

A decade marked by the exponential growth of information pushed the majority of cloud spenders towards a common goal: to reduce the cost of the cloud.

‍

Pathological spending waste arising from resource over-provisioning can be mitigated by many existing tools, but how much do these tools really penetrate the opportunity to reduce costs, when resource-intensive tasks – as common in big data and AI – are increasingly more widespread?

‍

We founded XONAI to address the fundamental issues that really increase costs for many organizations that rely on cloud computing: the need to scale resources for data and AI-driven products. To keep up with business growth, these products scale with more machines added to the resource pool, allowing them to maintain the timing demands of data preparation tasks and the latency constraints of AI servicing that are imperative to keep end-users engaged.

‍

The organizations we engage with are pushing the limits of big data analytics and its scale requirements, and we are helping them free up their time for more value-adding endeavors, rather than manual infrastructure optimization.

‍

We pioneered a solution to seamlessly accelerate data and AI processing, and we use the latest developments in compiler technologies to achieve it.

‍

Now, we are scaling our solution stack from accelerating big data processing towards incorporating integrations with AI frameworks, as end-to-end optimization of big data analytics workloads is more crucial than ever, considering the digital transformation that is taking place today in the new era of AI.

‍

We are tackling this challenge bottom-up – eliminating fundamental performance bottlenecks rippling across the entire stack and ultimately driving resource allocation.

‍

At the most granular level, the root cause of these bottlenecks are data movements at the boundaries of individual computations. We mitigate these with a purpose-built domain-specific language that acts as a single source of truth for expressing such computations – and combines them into optimized kernels prior to execution.

OUR APPROACH

Our approach is non-invasive by design, accelerating data processing on existing cloud hardware without modifying any original execution planning logic. It is even compatible with modified Spark runtimes such as Amazon EMR and Databricks. Performance is never lost as we always compound on top of existing platforms.

‍

Organizations can trust XONAI as a drop-in solution that reliably accelerates existing data pipelines without any code changes – even if these are already highly tuned for the environment they are being deployed, which they often are.

‍

Our benchmark results are consistent across the board, as we deliver cost reduction in every major big data platform for the cloud. Additionally, our solution does not rely on any particular hardware type or set up. It “just works” by reducing task execution time in a variety of hardware (e.g. Intel, AMD, AWS Graviton processors) and without any caveats, very often allowing moving to cheaper instances without any drawbacks as a result of more optimal memory and I/O resource management.

‍

Despite our incredible achievements this early on, we are barely scratching the surface of what we want to realize in the long run.

‍

Our mission is too fundamental to be solved in one go – and we want to do it right – by incrementally validating our novel approach in production-grade systems, and moving towards unifying data and AI optimization for organizations that want to realize greater value from their products.

MOVING FORWARD

We’re incredibly proud that London-based Kadmos Capital, together with Adara Ventures, Deep Science Ventures, Nauta Capital, Notion Capital, and notable angels Martin Gould (CEO and Founder of This One), Mehdi Ghissassi (Director of Product Management for Google DeepMind), among others, believe in our mission to optimize data infrastructure and participated in our Seed round. The total of $3.5 million in seed funding will allow us to scale our team and ensure enduring customer growth and support. We’re also thrilled to have Ashraf Lotfi and Graham York as new additions to our board – two incredible advisors bringing a wealth of experience in entrepreneurial and technological computing sectors to help shape the future of our company.

‍

At XONAI, we are committed to building the best-in-class solution for helping organizations to accelerate time to value and reduce infrastructure operational costs. We are scaling our amazing team to bring world-class support to our customers and expand our novel solution to optimize data and AI end-to-end and realize the universal compute fabric. If our mission resonates with you, have a look at our open roles or learn more about our company by subscribing to our newsletter at xonai.io.

Stay tuned for the many exciting updates to come! 🚀