Zeta Introduces AthenaOS, an Enterprise Intelligence Operating System

Zeta Global today at Zeta Live 2026 introduced AthenaOS, an enterprise intelligence operating system built on the Zeta Data Cloud and combining enterprise knowledge with Zeta's intelligence about who will buy, what they want, and what they will do next. It connects that combined intelligence with Zeta's intelligent AI infrastructure and then dynamically builds around objectives, helping teams identify growth opportunities, take action, and measure results.

AthenaOS starts with what users want to achieve. Users tell Athena by Zeta, the conversational interface, their objective, and AthenaOS assembles the intelligence, applications, and actions needed to achieve it. It then operates through a continuous loop, using each outcome to improve the next decision.

"For decades, software forced people to work the way applications were designed, but AthenaOS changes that," said David Steinberg, co-founder, chairman, and CEO of Zeta Global, in a statement. "AthenaOS marks an expansion of Zeta's intelligent AI infrastructure beyond marketing. It starts with what you want to achieve, brings together what your business knows with Zeta’s intelligence about the customer, and organizes the work around the outcome. That's the shift from navigating software to directing outcomes and how intelligence becomes growth."

AthenaOS introduces the following new capabilities:

• Intelligence organized around an objective: AthenaOS dynamically brings together the enterprise context, Zeta intelligence, and actions needed to help users achieve a specific business outcome.
• A continuous know, decide, act and learn loop: Every decision, action, and result that AthenaOS delivers strengthens the next one, helping improve the quality and speed of future decisions.
• Shared enterprise context: AthenaOS creates a shared context framework across enterprise knowledge, customer and market signals, and partner technologies such as Palantir Foundry and OpenAI, allowing context, memory, and intelligence to move with the work rather than remain trapped inside individual systems. This preserves each enterprise’s data sovereignty and control over how its information is used.

Zeta also introduced Athena MCP, its headless solution, which enables companies to bring Zeta's intelligence into their own applications, agents, and workflows. Teams can work directly with Athena, or apply its intelligence within systems like ChatGPT, Claude, Gemini, and other compatible enterprise AI environments.

"Disconnected tools cannot create connected outcomes, and adding more AI does not solve that problem on its own," said Christian Monberg, chief technology officer of Zeta Global, in a statement. "AthenaOS connects business context, people, applications, and AI agents around an objective. With Athena MCP, that intelligence can operate within the systems where teams already work and help them create the applications and workflows they need on demand, tailored to their businesses. The result is software that adapts to the work, with the shared context and controls enterprises need to operate at scale."

Underpinning AthenaOS is an enterprise-grade foundation of identity, governance, permissions, security and transparency. This foundation is designed to give enterprises greater control over how intelligence is applied across their data, applications and workflows.

Zeta also unveiled the Athena Inference Model (AIM), a collection of models developed with Fireworks using NVIDIA Nemotron for specific AthenaOS use cases. These models are fine-tuned to understand specialized skills like domain language and follow defined brand voice, formatting, and output requirements.

AIM gives Zeta unique insight into the intelligence workloads businesses need to run, creating the potential to optimize both the models and the computing beneath them.

"The Athena Inference Model is critical in how we bring our intelligent AI infrastructure to life and advance data and AI sovereignty," Steinberg said. "Models are increasingly available, but context is owned. AIM helps organizations harness their most strategic assets securely and effectively, connecting their data and context to AI in a way that improves how decisions are made and work gets done. That is the advantage we are building for our customers."

"General-purpose models provide the foundation, but open models unlock the proprietary data and specialization that drive the next leap in performance," said Nate Yohannes, president of data and AI Lab and global head of research and development at Zeta Global, in a statement. "AIM give AthenaOS a more efficient way to understand enterprise context, apply specialized skills and deliver more useful recommendations across the business."

"At Fireworks, we believe enterprises require specialized intelligence because data sensitivity cannot be an afterthought," said Benny Chen, co-founder of Fireworks, in a statement. "By combining Zeta's proprietary intelligence with Fireworks' infrastructure for training, tuning, and serving open models, AIM are more accurate, more efficient, and tailored to Zeta's specific business needs."