Ema, a startup that deploys teams of AI agents it calls “AI employees” to automate enterprise workflows, has raised $77 million in a Series B round led by Bengaluru-based venture firm Creaegis, bringing total funding to $140 million and more than quadrupling its valuation since its previous round in 2024, TechCrunch reported.
Key points
- Creaegis led the $77 million Series B; existing investors Accel, Section 32 and Prosus all increased their stakes.
- The financing was structured solely as primary equity, excluding any debt or secondary share sales, the company confirmed.
- Revenue has grown 50-fold over two years; revenue bookings exceed $150 million, though the figure includes multiyear contract value rather than annual recurring revenue.
- Net dollar retention is around 180 per cent, with more than 90 per cent of customers expanding beyond their initial use case.
- Ema has more than 50 active enterprise deals, over 1 million active enterprise users, and customers including NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro and Microsoft.
Series B led by Creaegis with existing investors doubling down
The financing was led by Creaegis, with every major existing investor participating at a higher level than before. Accel, Section 32 and Prosus each added to their positions, a decision the company attributes to the investors’ direct visibility into customer usage, renewals and expansion. The round was structured as primary equity only, with no debt or secondary components, Ema confirmed to TechCrunch. The startup declined to disclose its latest valuation.
Creaegis managing partner and chief investment officer Prakash Parthasarathy said Ema has become one of the quickest enterprise-ready AI firms, with proven solutions that span customer engagement, commercial activities and vertical use cases on a unified foundation.
Revenue growth and customer expansion metrics
Ema said its revenue has grown 50-fold over the past two years. Revenue bookings have topped $150 million, though chief executive Surojit Chatterjee clarified that the figure reflects the aggregate value of multiyear agreements rather than annual recurring revenue. He declined to disclose the current annualised revenue run rate.
The startup has more than 50 active enterprise deals and over 1 million active enterprise users, with more than 5 million actions and queries handled to date. Chatterjee told TechCrunch that more than 90 per cent of customers have expanded beyond their initial use case, some deploying the technology across dozens of workflows. Net dollar retention stands at around 180 per cent, indicating existing customers are substantially increasing spend over time.
Case studies cited by the company illustrate the scale of deployment. At a top-tier global systems integrator, Ema deploys a support tool for more than 240,000 personnel across 65 nations, automating over 100 processes and managing approximately 2.9 million queries annually. Response times have fallen from days to seconds, employee satisfaction has risen 20 per cent, around 60 per cent of tickets have been removed, and the people operations function operates about 50 per cent leaner. At a top-five global business process outsourcing firm, Ema’s AI employees process over one million IT service management tickets annually and manage more than one million calls in over 15 languages with up to 95 per cent accuracy, with under 10 per cent of interactions escalated to a human agent. For a $50 billion global conglomerate, the platform went from concept to live production in four weeks, connecting to over 20 systems of record, supporting more than 40,000 employees, and delivering a 70 per cent improvement in efficiency alongside a 30 per cent drop in ticket volumes.
Pricing tied to outcomes, not seats or tokens
Ema’s pricing model is linked to the completion of tasks and business outcomes rather than software seats or AI token consumption. Chatterjee said the company maintains gross margins of close to 80 per cent, with margins improving over time as its AI systems learn from deployments and require less human support. The platform merges AI reasoning with enterprise-grade safeguards for oversight, protection and regulatory adherence, and links to over 250 business applications and core systems.
Chatterjee told TechCrunch that Ema’s software can draw on a library of more than 150 AI engines, including frontier and open-source variants, while the startup concentrates on the subject-matter expertise, connections and coordination required to automate business processes from start to finish. He does not see frontier AI labs such as Anthropic and OpenAI as direct competitors, stating that progress in frontier models is beneficial to Ema. The company’s approach, he said, is to first wrap around an enterprise’s existing applications before customers can lessen reliance on some of these offerings and, in some instances, substitute them entirely.
Capital directed at go-to-market expansion
Much of the new capital will go toward expanding go-to-market operations, particularly sales and marketing, after the company spent its first years largely building the product. Multiple senior leaders have already been hired. Ema is headquartered in Mountain View, California, and was founded in 2023 by Chatterjee, a former Google and Coinbase executive, and Souvik Sen, a former Okta executive.
