AI has an Insatiable Appetite for Two Things: Compute and Context: That’s Why We Invested in Craxel

Barely a week goes by in the AI space without news of another ambitious project for an AI factory with a massive data center to meet the needs of model training, agent development, and inference at scale. Such reports generally include mention of the need for gigawatts of power and deployment of tens of thousands of the latest GPU chips. Costs are often cited too - often tens of billions of dollars.

While acknowledging the sheer scale of such developments and the commitment both in terms of the build-out effort and financial backing required before even the first lines of AI code can be executed, at Iona Star we tend to view algorithmic innovation as the key to feeding AI. 

Throwing hardware at the problem has long been a lazy response to software performance issues, and we believe it is a costly approach that has definite limitations. We are much more excited by smart, high-performance software built on cutting-edge algorithms that deliver accurate, trusted, and responsive AI with a more affordable cost base. Bringing them to market has informed our investment thesis.

Today, we are announcing our investment in Craxel, builder of the patented Black Forest Knowledge Infrastructure for AI-powered decision making. Based in Fairfax, Virginia, Craxel has to date focused on the demanding defense, aerospace, and financial services verticals, though Black Forest can support many applications in verticals that plan to implement enterprise AI applications for global customer bases.

“Investing in Craxel and its world class team was a no-brainer for us since we share a common AI worldview that true value is delivered through algorithmic innovation informed by deep real-world experience of deploying the most demanding, data intensive applications,” said Gerry Buggy, partner at Iona Star

To learn more about Craxel and Black Forest, we asked Founder and CEO David Enga, to answer some questions:

What is your professional background and how did you come to found Craxel?

I spent over 25 years designing mission-critical data systems for the U.S. defense and aerospace sectors, where we kept running into the same core problem: how to organize and query massive, fast-moving data efficiently and securely. The tools available couldn’t scale without extremely inefficient brute-force computation.

A digital library is the perfect analogy that describes the problem. Information is coming in so fast that the digital librarian can’t put the “data” on the shelves and update the “card catalog” fast enough. The information of today is also too complex for a simple card catalog approach to organizing it. Most people are unaware of the extent of this problem, but it is the root cause of today’s data challenges.  It gets worse every single day. Enterprises are rapidly becoming aware of this fundamental problem as they look to feed AI context.

So, we went deep into one of the hardest unsolved problems in computer science: how to organize data for fast, scalable, ad hoc query. While many researchers were chasing in vain for a scalable perfect order-preserving hash, we found a different path. We discovered a hierarchy-preserving probabilistic hash that provides an exponential advancement, a constant-time algorithm that organizes multi-dimensional data at line speed.

That algorithmic innovation, which we call O(1), became the foundation of Black Forest. It not only enables efficient ingest and instant query at massive scale but also supports this for fully encrypted data with rich access patterns: range queries, time series, vector search, spatial analysis, partial match, and fully dynamic and transactional knowledge graphs.

That’s the breakthrough. It solves problems legacy systems were never built to handle and it changes what’s possible for delivering context to AI.

How did Black Forest come about? What business drivers are you addressing?

Black Forest came out of hard challenges from the real world. We saw firsthand how large organizations, especially in defense and intelligence, were buried in data that users couldn’t even find or access quickly enough. The tools on the market couldn’t scale without resorting to brute force search methods, couldn’t secure data without compromises, and couldn’t deliver insight in time to act.

If legacy technology couldn’t rapidly deliver information and context for human users, how will it be able to for AI? Our discovery fundamentally changes the paradigm so that fully contextualized information can be rapidly delivered to both human users and algorithms.

Black Forest gives organizations instant access to fully connected data, so they can detect, decide, and respond in real time. It cuts down infrastructure costs, slashes compute requirements, and rapidly delivers exquisite context from enterprise data to AI.

It wasn’t about making yesterday’s systems better. It was about building what the future actually needs.

How would you characterize Black Forest’s key technical capabilities and why is it ideally suited to power AI applications?

AI doesn’t just need data. It needs fast, connected, contextualized data. That’s what Black Forest delivers. At the core is our O(1) indexing algorithm which organizes vectors, semantic triples, timelines, geospatial data, and more into massive multidimensional knowledge graphs, at line speed.

Black Forest isn’t built on legacy technology.  We built it from scratch in a whole new way.  It’s a fundamentally different engine, built to eliminate the compute and latency bottlenecks that cripple AI at scale. With Black Forest, triples and vectors can be indexed as they arrive and queried instantly. That means better performance for things like retrieval-augmented generation, real-time decision agents, and structured and unstructured search across massive knowledge graphs.

It also means you don’t need to choose between performance, cost, and security. Our architecture supports fully encrypted data with zero-trust access controls and high-speed queries.

In a world chasing more hardware and brute force, we chose better algorithms. That’s what enables faster, more accurate AI, with lower cost, stronger security, and unparalleled scale.

Is it possible to highlight some application use cases for Black Forest?

Black Forest was built to solve complex, high-scale problems where traditional data systems fall short. Here are a couple of examples:

In a military setting, imagine trying to track and intercept hypersonic missiles across thousands of fast-moving satellites. The data is high velocity, multidimensional, and spread across domains. Legacy systems rely on pre-processing and batch analysis, which are too slow to respond in real time. With Black Forest, telemetry, event, and threat data can be indexed as it arrives and fused into real-time, query-able knowledge graphs, so analysts and AI agents can detect, correlate, and act immediately.

In crypto surveillance, the problem is following the money through a constantly shifting web of transactions, wallets, exchanges, and identities and often across jurisdictions and chains. Black Forest turns these events into dynamic graphs that reveal hidden relationships and time-based patterns. Investigators can spot anomalies, trace flows, and flag fraud or illicit finance in seconds, not days, without having to scan entire blockchains every time.

That’s the power of instant access to fully connected data at any scale.

Why did you accept the investment from Iona Star, and how do you expect to work with the team here?

We were looking for partners who understood the scale of the problem we’re solving and the long-term value of real algorithmic innovation. Iona Star stood out immediately. They grasped the strategic importance of what we’ve built and why it matters, not just for AI, but for national security, critical infrastructure, and enterprise data at large.

What impressed me most was their technical curiosity and the way they think about durable advantage. Craxel has a deep moat, mathematics, patents, and software built from scratch, and Iona Star understood that clearly. Iona Star has a strong reputation for backing companies solving foundational, high-impact problems, and they bring experience and perspective that will support us as we scale.

We expect the relationship to be collaborative. They’ve already begun helping us expand our reach, opening doors to new conversations with other investors, partners, and decision-makers who care about the future of AI and data infrastructure.

Looking out to the future, we’re focused on scaling with the right customers; those who understand that algorithmic innovation is the only sustainable path forward in data and AI. We’re not chasing hype cycles or building around legacy constraints. We’re building the foundation that AI fundamentally requires.