Ema raises $77M to deploy AI employees that replace SaaS seats with outcome-based pricing

Ema, a startup building "AI employees" that orchestrate multiple agents to automate enterprise processes across HR, IT, and finance, has raised a $77 million Series B led by Creaegis with participation from Accel, Section 32, and Prosus, bringing total funding to $140 million and roughly quadrupling its prior valuation. The company reports more than 50 active enterprise deals, over 1 million active users, and revenue bookings surpassing $150 million with net retention around 180%. For builders, the most important signal is Ema's pricing model: it charges based on task completion and business outcomes rather than software seats or AI token usage, a structure that could reshape how enterprise automation is sold and delivered.

What the Series B means for enterprise automation

The round was led by Bengaluru-based Creaegis, with all existing major investors increasing their stakes. Ema's customers include NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro, and Microsoft. The company says it has processed more than 5 million actions and queries, and revenue has grown 50-fold over two years. CEO Surojit Chatterjee told TechCrunch that more than 90% of customers have expanded beyond their initial use case, with some deploying the technology across dozens of workflows. The bookings figure includes multiyear contracts, not annual recurring revenue, and Ema declined to disclose its current ARR.

Why outcome-based pricing matters for AI builders

Ema's pricing model is a direct challenge to the per-seat SaaS model that dominates enterprise software. Instead of charging per user or per API token, Ema ties revenue to task completion and business outcomes. Chatterjee said the company maintains gross margins close to 80% because its AI systems learn from deployments and require less human support over time. For builders designing AI products for enterprise, this model offers an alternative to usage-based or seat-based pricing, aligning vendor incentives with customer results. It also puts pressure on traditional SaaS vendors to justify seat-based pricing when AI agents can replace multiple human workflows.

How Ema's orchestration approach differs from frontier AI labs

Ema does not build its own foundation models. Instead, it draws on more than 150 models, including frontier and open-source models, and focuses on domain knowledge, integrations, and orchestration to automate end-to-end business processes. Chatterjee explicitly said frontier model progress benefits Ema, and he does not see OpenAI or Anthropic as direct competitors. This is a key distinction for builders: Ema's value lies in the agent orchestration layer and enterprise integrations, not in model capability. The platform first "wraps" around existing enterprise applications before customers can reduce dependence on some of those products, and in some cases replace them entirely.

Expansion plans and what's next

Ema plans to use the new capital to expand go-to-market operations, particularly sales and marketing, after spending its first years largely building the product. The company has grown to nearly 200 employees with offices in Bengaluru, London, and Vancouver. It now plans to expand into Asia-Pacific, South America, and parts of the Middle East over the next year. Chatterjee also noted that AI can take over some implementation, integration, and consulting work traditionally done by IT services firms, and that several services companies are already working with Ema to change their own business models.

Caveats to keep in mind

All metrics cited are company-reported and have not been independently verified. The $150 million bookings figure includes multiyear contracts, not annual recurring revenue, and Ema declined to disclose its current run rate. Claims about replacing large SaaS applications are based on CEO statements and customer anecdotes, not independent audits. The net retention rate of 180% reflects expansion within existing customers but may not be representative of broader market dynamics. Builders should treat these numbers as directional signals rather than verified benchmarks.