3 Years of Building Ema

Three years ago today, we started Ema with a conviction that felt obvious to us but contrarian to most then: the future of enterprise work wouldn't be about better software—it would be about AI employees working alongside humans.

In February 2023, this wasn't a popular take. "Agents" didn’t exist as a term outside of research projects. The idea of agentic AI systems autonomously handling end-to-end enterprise workflows sounded like science fiction.

We saw something different and decided to bet on it. Knowledge work is fundamentally composed of decisions, not just tasks. If you could build systems that reason, learn, and act within defined parameters—that make judgment calls, not just follow scripts—you could transform what software was capable of.

So we started building.

The Early Days

The first year was humbling. We weren't building a better CRM or a smarter analytics tool. We were building an entirely new category—a horizontal agentic operating system for the enterprise. You would be able to conversationally build and deploy an AI Employee to 10x productivity in any function of the enterprise.

That meant convincing large companies to trust AI with their most critical workflows: customer support, financial operations, legal processes, sales execution.

The skepticism was real. "This sounds great in theory, but can it actually work in production?" Every prospect asked some version of that question. Fair question. We didn't have proof yet—just conviction.

The only way to answer it was to build something that actually worked. Not a demo. Not a prototype. A production-grade system with the architectural trust, transparency, and reliability that enterprises demanded.

The Compound Effect of Great People

What made those early days possible—and what's made the last three years possible—is the team we've built.

I'm often asked what I'm most proud of at Ema. The answer isn't our technology or customer wins, though I'm proud of those too.

It's the team. We assembled people who are world-class technically AND who deeply understand that we're transforming how real humans work.

People who obsess over edge cases.

Who build with empathy.

Who ship fast without cutting corners on trust.

Across three continents, different backgrounds, different experiences—united by a shared belief that this technology can make work more human, not less.

Grateful for our Investors, Advisors, Customers, Partners

We've been fortunate to have investors and advisors who got the vision early and stuck with us through the inevitable ups and downs of building something new. Who challenged us when we needed challenging. Who supported us when things were hard.

To our customers: you took a bet on us when agentic AI was unproven. You gave us honest feedback when things broke. You pushed us to think bigger about what was possible. The learnings from your deployments are in every line of code we write.

To our partners: you amplified our reach and deepened our impact. You helped us navigate complex enterprise environments. You co-created solutions that neither of us could have built alone.

Where we are now

The world has caught up to what we saw in 2022. Agentic AI went from contrarian to consensus. From science fiction to boardroom strategy.

But consensus brought noise. The term "agentic AI" became a buzzword. The market flooded with vaporware. And 2024-2025 brought a reckoning: 85-95% of enterprise AI initiatives failed to deliver value.

We stayed focused on the work. We deployed AI employees across sales, support, finance, legal, employee experience—building the horizontal agentic OS we'd envisioned. Not in controlled environments. In production. Processing millions of workflows. Learning from failures. Earning trust through architecture.

The result: we're now working with Fortune 500 companies and fast-growing enterprises who've moved beyond "Can this work?" to "How fast can we scale this?" We've become one of the handful of companies that separated promise from reality in 2025.

Where we’re going

Our mission hasn't changed: build a horizontal agentic operating system that enterprises trust with their most critical work.

But our understanding of what that means has deepened. It's not just about agentic AI that can execute tasks. It's about systems that expand what organizations are capable of.

That enable strategies that were previously cost-prohibitive.

That free humans to focus on judgment, creativity, and relationships while AI handles the repetitive, process-driven work.

It's about building technology that makes work more human, not less.

It's about doing this at a pace that matches the urgency of the opportunity, but with the discipline and architectural integrity that production systems demand.

Looking at the next three years, we believe Agentic Business Transformation will happen across large enterprises exponentially. The companies that figured this out in 2024-2025 are entering 2025-2026 with compounding advantages.

We're ready. Our technology is battle-tested. Our team is world-class. Our customers are scaling from functions to enterprise-wide deployments.

But I'm most excited about what we haven't built yet. The capabilities we'll unlock. The use cases we haven't even imagined yet. The ways agentic systems will evolve as we learn from millions more workflows.

Thank You

Three years feels both long and short. Long because of everything we've built and learned. Short because of how much we still want to do.

To everyone who's been part of this journey—team, investors, advisors, customers, partners—thank you. Here's to the next three years of building the future of work—one production deployment at a time. 🚀