OPAQUE Systems is a confidential AI platform unlocking sensitive data to accelerate AI into production securely. Pulse 2.0 interviewed OPAQUE Systems CEO Aaron Fulkerson to gain a deeper understanding of the company.
Aaron Fulkerson’s Background
What is Aaron Fulkerson’s background? Fulkerson said:
“I’ve spent my career applying emerging technologies to drive human progress—always with a focus on responsibility, openness, and impact. From open-source platforms used by hundreds of millions to billion-dollar AI product lines, I’ve prioritized open standards, data privacy, and scalable innovation.”
“My journey started with a simple goal: use technology to help people transcend their circumstances. That meant building tech community centers with internet access in underserved parts of North Carolina, researching distributed systems at Microsoft, and launching open-source tools that put power in the hands of others. Even as I took on broader leadership roles, the mission remained the same: build products that solve macro challenges and respect the individual.”
“Now at OPAQUE, we’re tackling what may be the biggest challenge yet—how to safely and ethically use the world’s most sensitive data to power AI. Our platform enables enterprises to run AI directly on encrypted data—no exposure, no trade-offs—and with verifiable guarantees of privacy and sovereignty. It’s a breakthrough that allows companies to move faster, stay compliant, and finally unlock the value of data they’ve never been able to touch. That’s what excites me most—and what we’re just beginning to scale.”
Formation Of The Company
How did the idea for the company come together, and what are your primary responsibilities? Fulkerson shared:
“Every CIO I talk to is facing the same problem: how to use their most valuable data in AI without violating privacy or losing control. That’s why we built OPAQUE.
Born out of UC Berkeley’s RISELab—the same team that created Databricks and Apache Spark—OPAQUE was founded by pioneers in distributed systems, cryptography, and confidential computing:
— Ion Stoica, creator of Apache Spark, Ray, and founder of Databricks and Anyscale; — Raluca Ada Popa, Grace Hopper Award winner for breakthroughs in data privacy; — Rishabh Poddar and Chester Leung, leaders in secure multiparty collaboration.
OPAQUE enables enterprises to run AI workloads on encrypted data at cloud scale, with verifiable guarantees of privacy and data sovereignty. We support the tools teams already use—Python, Spark, ML libraries—and enable secure LLM and RAG deployments on data that was previously out of reach.”
“I stay deeply involved with Product because understanding our customers’ challenges is the only way to build the right solution. I spend much of my time working directly with CIOs, CTOs, CISOs, and Heads of Engineering to ensure OPAQUE helps them unlock sensitive data for AI—without trade-offs. That includes LLMs, RAG, and other agentic workflows that demand strong and security–verifiable zero-trust. I also work closely with strategic partners like Microsoft Azure, NVIDIA, and Accenture to expand our reach and strengthen the Confidential AI ecosystem.”
Favorite Memory
What has been your favorite memory working for the company so far? Fulkerson reflected:
“My favorite memories are those where I see our customers realize they can finally do what they thought was impossible with AI—without giving up control of their most sensitive data.”
“One standout moment was hearing Bill McDermott on MSNBC explain how ServiceNow transformed their business using AI agents. He described how a workflow that used to take 8 days now takes just 4 seconds. That leap was made possible by OPAQUE’s confidential agents for RAG—and it was incredibly fulfilling to see our impact show up at that scale.”
“But my top career highlight has been our work with the European Union to combat organized crime operating across EU borders. We enabled AI to analyze sensitive threat data while preserving citizen privacy—something that was previously off-limits. It’s a powerful example of how you can drive innovation and protect fundamental rights at the same time.”
“Those lightbulb moments—where customers realize they can solve their hardest problems without compromise—are what drive me. It’s why we built OPAQUE.”
Core Products
What are the company’s core products and features? Fulkerson explained:
“OPAQUE started as a confidential computing version of Spark—focused on secure analytics and ML—but it has since evolved into a full Confidential AI platform for deploying large language models (LLMs) on sensitive data.
At the core of the platform are confidential agents: policy-bound AI workflows that provide verifiable guarantees of privacy, security, and data sovereignty. These agents support Retrieval-Augmented Generation (RAG), enabling organizations to use their most sensitive data in AI—like customer records, financial data, or intellectual property without exposure.
These verifiable guarantees are what make adoption possible. They ensure that every query, output, and data access is bound by policy, cryptographically verifiable, and safe to run in production. That’s the key to unlocking enterprise AI at scale.”
Challenges Faced
What challenges has Fulkerson and the team faced in building the company? Fulkerson acknowledged:
“OPAQUE enables organizations to accelerate their AI projects by 4-5x while securing sensitive data. But one of our biggest challenges has been awareness—most technology leaders still don’t know this is even possible. They’re waking up to the enormous risks AI introduces, but the solution—Confidential AI—is still new and relatively unknown.”
“What’s helped is the growing validation from major tech companies. Apple’s Private Cloud Compute, Meta’s Private Processing for WhatsApp, and production deployments from Microsoft, Google, Amazon, and TikTok—all signal that Confidential AI is real, proven, and already in production. These companies are using it to protect their most sensitive data on their infrastructure.
This reminds me of the early days of the internet. There was a time when people logged into bank accounts or shopped online over open public Wi-Fi using HTTP. It sounds insane now—but at the time, it was normal. Then HTTPS became the standard, encrypting communication by default. We’re seeing the same inflection point with AI. In 2–3 years, Confidential AI will be the default—because it’s the only way to safely run AI on the data that actually matters.”
“Additionally, our annual Confidential Computing Summit is accelerating awareness (www.ConfidentialComputingSummit). By convening technology leaders and their customers from across industries, we’re shining a light on what’s possible with Confidential AI—and helping organizations see how to move from risk and regulation to innovation and adoption.”
Significant Milestones
What have been some of the company’s most significant milestones? Fulkerson cited:
“In 2022, we secured a $22 million Series A, which helped us bring Confidential AI into production with some of the world’s most data-sensitive organizations. Since then, we’ve raised $31.6 million and seen our customers—like ServiceNow, Encore Capital, the European Union, and RiskStream Collaborative—use OPAQUE to unlock sensitive data and drive real AI impact.”
“What’s been most exciting is witnessing the growth of the Confidential AI category itself. It’s gone from niche to necessary—both in the enterprise and in the broader market. The number of competitors entering this space has exploded. We see new startups every week, which tells us one thing: organizations everywhere are looking for a better, safer way to adopt AI.”
“Our third annual Confidential Computing Summit is happening June 16–18 in San Francisco, and it’s our biggest yet. We’re bringing together every major cloud provider, confidential AI users across all industries, and the fastest-growing startups in the space, like CrewAI, LangChain, and Bloomfilter. The Summit shows that Confidential AI is no longer theoretical—it’s being deployed, trusted, and scaled by the people who matter most: our customers. And everything we do is focused on helping them adopt AI responsibly and effectively.”
Differentiation From The Competition
What differentiates the company from its competition? Fulkerson emphasized:
“We created the category of Confidential AI. While others are just entering the space, we’ve been building and scaling with enterprise customers from day one. Our founders—brilliant researchers from UC Berkeley’s RISELab—gave us a deep technology advantage with years of innovation in distributed systems, secure computing, and AI.
That foundation allowed us to move fast, and more importantly, to lead. While new competitors are popping up weekly, OPAQUE is already powering Confidential AI in production at global scale—with customers like ServiceNow, the European Union, and Encore Capital. Others are validating the category; we’re defining it.”
Future Company Goals
What are some of the company’s future goals? Fulkerson explained:
“I’m excited about where OPAQUE is headed, because our customers are just getting started. Every week, we see more organizations adopt Confidential AI to solve real problems—safely, ethically, and without compromise. Our third annual Confidential Computing Summit this June brings that to life: leaders from Microsoft, NVIDIA, Accenture, and the hottest AI startups will be joined by enterprise buyers using OPAQUE in production. It’s a clear signal that this category is accelerating.
Looking ahead, our focus is simple: help more customers adopt AI responsibly. That means giving them verifiable guarantees—not just of privacy and sovereignty, but also that their data, models, and responses can’t be exfiltrated or misused. As more companies realize that basic encryption isn’t enough, they’re turning to OPAQUE to future-proof their AI strategy—and we’re here to help them get there.”
