Funding Brief｜Ema Raises $77M Series B as 'AI Employees' Start Eating Enterprise Software Budgets

🌏 中文版

Funding Details

What the company does

Ema builds enterprise agent orchestration — rather than selling a single-task AI tool, it sells a coordinated team of AI agents it calls "AI employees" that take over multi-step business processes spanning HR, IT, and finance.

Its approach is to "wrap" a company's existing applications: Ema deploys a layer of agents around a customer's current SaaS systems, which reduces those systems to little more than databases, letting customers gradually cut back on — or eventually drop entirely — their dependence on the large SaaS products underneath. Behind the scenes, Ema can draw on more than 150 models, spanning both frontier and open-source options. Co-founder Surojit Chatterjee (a former Google and Coinbase executive) says the company focuses on domain knowledge, systems integration, and cross-application orchestration rather than competing with frontier labs directly — "progress in frontier models is actually very beneficial to us." Pricing is deliberately not tied to seats or token consumption; it's tied to completed tasks and business outcomes instead.

Ema now has more than 50 active enterprise customers and over a million active users, and has handled more than 5 million actions and queries. Its customers include NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro, and Microsoft. Revenue has grown 50x over the past two years, and cumulative contract bookings (including multi-year deals) have surpassed $150M. More than 90% of customers expand from their initial use case to dozens of workflows, net dollar retention sits around 180%, and gross margin remains close to 80%.

What this round signals

What it means for the agent ecosystem

Most of the new capital is going toward sales and marketing expansion — Ema spent its first few years mostly building the product, and this round is meant to push "AI employees" into more markets, particularly Asia-Pacific, South America, and the Middle East. That also means Ema isn't just selling software; it's going after the implementation, integration, and consulting work companies used to pay IT services firms to do. Several services firms are already partnering with Ema, in part because they recognize their own human-forward business model may not survive the next stage.

What investors are betting on

Creaegis led the round with Accel, Section 32, and Prosus all adding to their existing stakes — a bet on whether Ema's "AI employee" model can keep holding its ground as frontier labs push directly into the enterprise market (Anthropic has already shipped plugins for finance, engineering, and legal work, and OpenAI has stood up teams of forward-deployed engineers). A valuation that more than quadrupled, with the amount undisclosed, suggests investors are willing to pay a premium bet that Ema keeps its differentiation — cross-application process orchestration — rather than getting absorbed by native features from the frontier labs themselves.

Numbers worth watching

• Cumulative contract bookings passed $150M, but the company explicitly declined to disclose annualized recurring revenue (ARR) — multi-year contract value and ARR sit at very different scales, so the real annualized revenue figure is still worth watching
• Net dollar retention around 180% is well above the 110-130% range typical of enterprise software, meaning expansion within existing accounts — not new customer acquisition — is the primary growth engine
• Gross margin remains near 80% even under an "AI replaces labor-intensive services" business model, close to traditional SaaS margins — one of the more unusual numbers in this round's disclosures

Watchlist status

Ema is not yet tracked in the watchlist. Recommend adding it under section D10 (Enterprise AI Platforms), alongside Enhans, Moveworks, and Glean, with tracking focus on: cross-department "AI employee" orchestration, its strategy of wrapping existing SaaS systems, and whether net dollar retention and gross margin hold up as it scales.

Today's takeaway

Most enterprise agent startups get framed as eventual head-on competitors to frontier labs like Anthropic and OpenAI, but Ema's founder offers a different read: model progress actually helps the company, because Ema isn't betting on "whose model is smartest" — it's betting on "who can plug a bunch of frontier models into real enterprise systems and actually finish a whole cross-department process." That suggests the enterprise agent market is splitting into two layers — model capability, and process orchestration on top of it — and the second layer isn't necessarily subordinate to the first.

References

• Ema raises $77M as AI starts eating into enterprise software and services
• Ema Raises $77M Series B to Put AI Employees to Work Across the Enterprise
• Ema secures $77m Series B to scale AI Employees