Ema raised $77M in a Series B round led by Bengaluru-based Creaegis, with existing backers Accel, Section 32, and Prosus adding to their positions. The financing brings the startup's total funding to $140M and more than quadruples the valuation it carried at its last round in 2024, though Ema declined to disclose the new mark. The round was entirely primary equity, with no debt or secondary component.
Founded in 2023 by former Google and Coinbase executive Surojit Chatterjee and ex-Okta executive Souvik Sen, Ema sells what it calls "AI employees" — orchestrated teams of AI agents that execute multi-step workflows across HR, IT, and finance systems. The pitch is that these agents don't just handle a single task; they wrap around a customer's existing SaaS stack and, over time, reduce dependence on it.
The traction numbers are the point of the round. Ema has 50+ active enterprise deals, more than 1M active enterprise users, and has processed over 5M actions and queries. Revenue has grown 50-fold over the past two years, with revenue bookings — the total value of multiyear contracts — surpassing $150M. Chatterjee declined to disclose annualized run rate.
“Many of our customers are already on the way to replace [large SaaS applications] completely, removing dependency on them, because they are mostly becoming like a database.”— Surojit Chatterjee, Ema co-founder and CEO
Key facts
- 01Ema raised $77M in Series B led by Creaegis, bringing total funding to $140M and more than quadrupling its 2024 valuation.
- 02The startup has 50+ active enterprise deals, 1M+ active users, and has handled 5M+ actions and queries.
- 03Revenue grew 50-fold over the past two years, with bookings past $150M and net dollar retention around 180%.
- 04Customers include NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro, and Microsoft.
- 05Ema draws on 150+ models, prices by task completion rather than seats or tokens, and holds gross margins near 80%.
Customer expansion is the more telling metric. More than 90% of Ema's customers have moved beyond their initial use case, with some deploying the agents across dozens of workflows. Net dollar retention sits around 180%, meaning the average existing customer is spending nearly twice what they did a year ago. The named customer list — NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro, and Microsoft — leans heavily on IT services and consulting firms, the same category Ema is angling to disrupt.
Chatterjee does not frame frontier AI labs as competitors. Ema's platform can draw on 150+ models, mixing frontier and open-source options, and the company positions its moat as domain knowledge, integrations, and orchestration rather than the underlying reasoning. Both Anthropic and OpenAI have pushed deeper into enterprise deployment this year — Anthropic with Claude in financial and legal workflows, OpenAI with forward-deployed engineering teams — but Ema's argument is that better base models make its orchestration layer more valuable, not less.
“Progress in frontier models is actually very beneficial to us.”— Surojit Chatterjee, Ema co-founder and CEO
Pricing is structured to reinforce the outcome pitch. Ema does not charge per seat or per token; it charges on task completion and business outcomes. Gross margins sit near 80%, which Chatterjee attributes to declining human-support costs as the AI systems learn from repeated deployments. That margin profile, if it holds at scale, is closer to enterprise SaaS than to the services businesses Ema is targeting.
The services angle is where the story gets more consequential. A significant share of enterprise IT spend goes not to software licenses but to the implementation, integration, and consulting work that firms like Wipro, PwC, and KPMG perform around that software. Ema is arguing — and its customer list of those same firms suggests it is being taken seriously — that AI can absorb a meaningful portion of that work.
The services incumbents are, per Chatterjee, aware of the shift and adjusting their own models rather than resisting it.
Ema has grown to roughly 200 employees, headquartered in Mountain View with offices in Bengaluru, London, and Vancouver. Most of the new capital will go to sales and marketing, after several years focused on product build-out. The company plans to expand from its current U.S. and Europe base into Asia-Pacific, South America, and parts of the Middle East over the next year.
The counterweight to the story is durability. Net dollar retention of 180% and 50-fold revenue growth are the kind of numbers that arrive early in a category and compress as the market matures. Ema's ability to sit on top of shifting frontier models is an asset today but leaves it exposed if a frontier lab decides to build its own orchestration layer with tighter model integration. The 150-model breadth is a hedge; it is not a guarantee.
The broader read is that the agent layer is starting to command real enterprise budget, and the money is coming out of two pockets simultaneously: SaaS license spend and IT services fees. Ema's 180% net dollar retention and its customer roster of the world's largest consulting firms suggest that the shift is not theoretical anymore. The interesting question for 2026 is whether the frontier labs let independent orchestration companies own that layer, or move to collapse it into their own enterprise stacks — and Ema's quadrupled valuation is a bet on the former.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
