Darren Kimura, CEO e Presidente di AI Squared – Serie di interviste

Darren Kimura is an experienced entrepreneur, inventor, and investor, currently serving as President & CEO of AI Squared. With over 25 years of leadership across the technology and clean energy sectors, he has led companies such as Energy Industries Corporation, Sopogy, and LiveAction, and co-founded the venture firm Enerdigm Ventures. Known for inventing MicroCSP solar technology, Kimura has played a key role in scaling companies focused on edge computing, AI, and SaaS-based solutions.

AI al quadrato is a low-code platform designed to help enterprises integrate AI insights directly into the tools they already use—such as Salesforce, Slack, Google Sheets, and ServiceNow. By simplifying AI deployment and enabling real-time decision-making, the platform empowers teams across engineering, sales, and operations to unlock measurable business value. Serving clients in finance, supply chain, and government, AI Squared is accelerating enterprise AI adoption by making it seamless, contextual, and actionable.

You’ve led multiple companies across sectors—what inspired you to join and ultimately lead AI Squared?

I’ve always been drawn to solving hard, real-world problems that require deep technology understanding and flawless execution. I began my career in energy efficiency and renewables, capturing machine data at scale and helping people understand and act on it. Over time, that evolved into moving data processing closer to where the work happens from the edge, in secure systems, or inside enterprise workflows.

When I first encountered AI Squared, I immediately saw that they were tackling a very real and very urgent challenge: how to move AI from the lab into everyday use. Most companies are still focused on building models. AI Squared was focused on deployment.

What gave me even more conviction was their foundation. The company started inside the NSA and has spent the past six years earning trust across the Department of Defense, the Intelligence Community, the Space Community and with some of the largest enterprises on the Fortune 500 list through real projects and meaningful results that continue to grow year over year. That told me two things: first, they were serious about security, compliance, and impact. And second, they were a team I could trust.

How has your experience scaling ZEDEDA and working in venture capital shaped your approach to AI deployment at the enterprise level?

At ZEDEDA we worked at the far edge as an operating system inside the open-source community. That forced me to deeply understand the real-world constraints our customers were facing with digital transformation. There we had a world class engineering team from Cisco, Sun Microsystems, Juniper, Arista but also taught me how important it is to pair that technical excellence with a strong enterprise go-to-market motion.

From the venture side, I gained a different kind of insight. I saw how deep tech companies scale, slower at first, but faster as you built upon a strong foundation. It taught me to think in systems: market timing, product readiness, customer maturity, and capital strategy all have to line up. That’s the lens I brought into AI deployment at AI Squared. It’s not about shipping models, it’s about delivering outcomes that enterprises can trust, measure, and scale.

AI Squared emphasizes embedding AI into systems people already use. What does solving the “last mile” of AI actually look like in practice?

87% of AI projects never reach production because integration is slow, complex, and disconnected from end users.  AI Squared solves this by creating a national security-grade data integration pipeline that connects AI/ML models with data sources like Databricks or Snowflake and the systems these companies already use like CRMs, ERPs, and custom apps.

How does your platform help bridge the gap between data science teams and business users?

AI Squared simplifies the handoff from model development to production.  Data science teams can deploy models securely through our platform, while business users receive insights within tools they already use inside Salesforce, Teams, Slack or any other system of record applications.

You’re working with the NSA, U.S. Navy, and other federal agencies. How do priorities like compliance, transparency, and ethical use shape AI Squared’s product roadmap?

Ethics, compliance and transparency are the top considerations on our roadmap.  We know from working with federal agencies, we must prioritize traceability, auditability, and explainability. We build with zero-trust principles, support for secure deployment environments, and clear governance models.

What role do you believe discriminative AI plays in improving decision-making for federal agencies?

Discriminative AI is precise, interpretable, and efficient. It’s ideal for high-stakes environments where you need clear answers, yes or no, threat or no threat. In federal agencies, this enables faster triage, risk detection, and prioritization. It complements generative models by offering structured, validated outputs that support critical decision-making.

When it comes to national security, where do you see the greatest risk—and opportunity—for AI adoption?

The greatest risk is unmanaged autonomy, deploying AI systems without transparency, oversight, or fail-safes. The opportunity lies in accelerating situational awareness and decision velocity. AI can sift through vast data in real time to surface patterns humans might miss. But to realize that value, it must be trusted, traceable, and integrated with human-in-the-loop systems.

What are some of the most common mistakes companies make when deploying GenAI?

The biggest mistake is focusing on novelty over necessity. Too many companies deploy GenAI without a clear use case, governance plan, or success metric. That leads to pilot fatigue, unscalable prototypes, and trust issues. GenAI is powerful, but like any technology, it needs to be aligned with the problem, the user, and the workflow.

Can you share an example where AI Squared helped an enterprise avoid “shiny object syndrome” and instead deliver measurable impact?

One enterprise came in wanting to build a custom LLM to classify customer service tickets. We showed them how a simple, embedded classifier deployed through our platform could deliver 90% accuracy in their CRM in less than two weeks. No new interfaces, no custom training. That’s the power of solving the right problem with the right tool.

What emerging trends are you most excited about in enterprise AI over the next 3–5 years?

Tre spiccano:

1. LLMs that are truly multi-modal and focused on specific tasks
2. Agentic AI systems that take action, not just generate content
3. Standardized governance that enables AI adoption in regulated industries

These trends are driving AI toward being more usable, trustworthy, and embedded in day-to-day operations.

How is AI Squared evolving its platform to keep pace with rapidly advancing LLMs and model ecosystems?

We’re model-agnostic by design. Whether it’s an open-source model, a commercial LLM, or a proprietary discriminative model, our platform allows organizations to securely deploy and manage it within their existing workflows. We're focused on orchestration, version control, and governance, so the model stack can evolve without disrupting the user experience.

What does success look like for AI Squared in the next chapter under your leadership?

Success means becoming the default way enterprises and government agencies operationalize AI. It means reducing time-to-value from months to days, and making AI deployment as seamless and secure as launching a new feature. Under my leadership, we’re focused on scale, trust, and real-world outcomes, not just innovation for its own sake, but impact where it matters most.

Grazie per l'ottima intervista, i lettori che desiderano saperne di più dovrebbero visitare AI al quadrato.