Ema is betting that the software industry is ready to move away from per-seat or per-token pricing models in favor of charging directly for task completion and business outcomes. Following a $77 million Series B round that brings its total funding to $140 million, the company is testing whether enterprises will prefer paying for results rather than software licenses. By aligning revenue with the actual value delivered, Ema aims to challenge the standard SaaS model, which has long relied on predictable, recurring revenue tied to user counts rather than the work performed by AI agents.
This shift toward outcome-based pricing mirrors a broader trend in vertical agent economics, similar to what we observed with Sela’s recent $21 million raise for voice AI agents in the mortgage sector. Both companies are betting that enterprises will prioritize measurable results over traditional software subscriptions. Ema, founded in 2023 by Surojit Chatterjee (ex-Google, ex-Coinbase) and Souvik Sen (ex-Okta), has attracted backing from investors including Creaegis, Accel, Section 32, and Prosus. The company reports a roster of Fortune-scale customers, including NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro, and Microsoft, claiming over 50 active enterprise deals and more than 1 million active enterprise users.
Ema’s approach involves integrating its multi-agent systems — what it calls “AI employees” — with existing enterprise applications across HR, IT, and finance, using more than 150 frontier and open-source models to coordinate workflows. The goal is to eventually render those legacy tools redundant. As CEO Surojit Chatterjee puts it, “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.” This aggressive stance on replacing legacy software contrasts with the approach seen in Salesforce’s Claudeforce deal, where major incumbents are hedging by integrating specific AI models rather than abandoning their own ecosystems entirely. It is important to note that Ema’s replacement trajectory is the company’s stated positioning, not an independently validated market trend.
Financially, Ema reports a 50x revenue increase over two years, 180% net dollar retention, and gross margins close to 80%. More than 90% of customers have expanded beyond their initial use case, the company says. However, these figures require careful interpretation. The reported $150 million in bookings represents the total contract value of multi-year deals — typically two- to three-year contracts — rather than annual recurring revenue. Chatterjee declined to disclose the company’s current ARR. All metrics are company-sourced and have not been independently audited.
This lack of standard reporting highlights the broader measurement problem we have covered previously. Without agreed-upon industry metrics, it is difficult to verify whether these agents are truly replacing legacy software or simply sitting alongside it as an additional cost layer. Given that only 5-8% of GenAI pilots achieve measurable at-scale ROI, the gap between ambitious claims and verifiable performance remains wide.
Ema is also positioning itself to capture spend beyond software budgets by targeting the IT services sector. Chatterjee explains that services firms are actively collaborating with the company, effectively disrupting their own traditional business models. “A lot of the services companies are working with us,” Chatterjee told TechCrunch. “They are also dramatically changing or disrupting their own business models because they understand the human-forward model may not be the best model going forward.” By inserting itself into both software and services workflows, Ema is attempting to capture a larger share of the enterprise technology budget.
While Ema is expanding into new regions including APAC, South America, and parts of the Middle East, the company must still prove that its model can scale beyond its early-adopter base. Replacing the backbone of enterprise operations is a different proposition than wrapping around it, and the industry is still waiting to see whether these agents can deliver on that promise at a truly enterprise-wide scale.