Analysis
Ema, the Mountain View-based enterprise AI agent startup, raised a $77 million Series B led by Bengaluru-based Creaegis, with existing investors Accel, Section 32 and Prosus increasing their stakes, according to TechCrunch and SiliconANGLE. The round brings Ema's total funding to $140 million and more than quadruples its valuation from its last raise in 2024.
What Ema Actually Sells
Ema was founded in 2023 by Surojit Chatterjee, a former Google and Coinbase executive, and Souvik Sen, a former Okta executive. The company builds teams of AI agents -- it calls them 'AI employees' -- that handle complete workflows inside corporate HR, IT and finance departments, rather than a single-purpose assistant bolted onto existing software. That framing puts Ema in more direct competition with the enterprise software and outsourced-services vendors whose work its agents are meant to replace than with other AI copilots.
“## What Ema Actually Sells Ema was founded in 2023 by Surojit Chatterjee, a former Google and Coinbase executive, and Souvik Sen, a former Okta executive.”
The Numbers Behind The Markup
Ema discloses unusually specific metrics for a Series B company: more than 50 enterprise deals, over 1 million active users, roughly 180% net dollar retention, 50-fold revenue growth over two years, and more than $150 million in bookings. Net dollar retention near 180% means existing customers are expanding their spend with Ema fast enough to nearly double revenue from the same customer base within roughly a year, a figure well above the 100-120% range considered healthy for enterprise SaaS -- though bookings and revenue are not the same thing, and Ema has not disclosed its actual revenue figure alongside the bookings number.
The Competitive Field
Ema competes for enterprise AI-agent budget against a widening field: Sierra, Glean and a growing set of vertical AI-agent startups on one side, and incumbent enterprise software vendors like Workday, ServiceNow and SAP adding their own agentic features on the other. Its differentiation is breadth -- covering HR, IT and finance workflows under one platform rather than a single department -- which is a harder integration problem but a larger total addressable market if it works.
What Founders And GPs Should Watch
The real test for Ema, as for most AI-agent companies claiming to replace headcount or outsourced services, is whether the reported 50x revenue growth holds up as it moves from early enterprise adopters willing to experiment to more risk-averse buyers who need to see multi-year reliability before replacing existing HR and finance workflows. A round this size, with three existing investors all increasing their positions rather than a new lead setting the price alone, suggests Ema's current backers have more conviction in the company's trajectory than the market broadly has had time to price in.
Chatterjee's background running product at Google and later serving as Coinbase's chief product officer, paired with Sen's identity and access-management experience at Okta, shapes Ema's specific bet: that the hardest part of deploying AI agents at enterprise scale isn't the underlying model, it's the permissioning, workflow integration and trust infrastructure needed for an agent to safely act inside a company's HR, finance and IT systems. That's a different emphasis than pure-model AI-agent startups, and it's closer to the problem Okta itself built a business solving for human employees.
Creaegis leading from Bengaluru rather than a US or European fund is also worth noting -- it puts an India-based growth investor in the lead position on a Series B for a Mountain View-headquartered company, reflecting how global the capital chasing enterprise AI agents has become this cycle, not just the founders and customers building in the category.