Ema raises $77M to automate enterprise work across HR, IT and finance

Ema has raised $77 million in a Series B round as the startup tries to move AI agents from isolated experiments into the business processes handled by enterprise software and IT services firms.

The financing, led by Bengaluru-based Creaegis with participation from Accel, Section 32, and Prosus, brings Ema’s total funding to $140 million, according to TechCrunch. The company said the round was entirely primary equity, with no debt or secondary transactions. Ema did not disclose its latest valuation, but said it had more than quadrupled since its previous round in 2024.

The funding arrives as enterprise buyers face a growing choice between adding AI capabilities to existing software and adopting systems that can perform work across several applications. Ema says it has more than 50 enterprise customers, including Google and Microsoft, but the company’s revenue, bookings, and usage figures are self-reported.

Deal and strategy behind Ema’s expansion

Founded in 2023 by former Google and Coinbase executive Surojit Chatterjee and former Okta executive Souvik Sen, Ema has built what it calls “AI employees.” These systems coordinate multiple AI agents to complete multi-step processes across a company’s existing applications.

That positioning is broader than a conventional coding assistant or chatbot. Ema is targeting workflows in HR, IT, and finance where employees and consultants typically move information between systems, apply business rules, and complete a sequence of actions. The company first operates alongside existing applications, Chatterjee told TechCrunch, with the longer-term possibility of reducing dependence on some software products.

Ema plans to use much of the new capital to expand sales and marketing. The startup spent its early years focused primarily on building the product and now has nearly 200 employees, with offices in Bengaluru, London, Vancouver, and Mountain View. It has mainly sold to customers in the United States and Europe and plans to pursue markets in Asia-Pacific, South America, and parts of the Middle East.

The company is also presenting its product as an alternative to some implementation and consulting work around enterprise software. Chatterjee told TechCrunch that services firms are working with Ema while reassessing models built around human-led integration and consulting.

What Ema says about its product and traction

Ema says its platform can use more than 150 AI models, including frontier and open-source models. The startup’s claimed differentiation is therefore not ownership of a foundation model, but the integrations, domain knowledge, and orchestration required to complete an end-to-end business process.

TechCrunch reported that Ema claims more than 1 million active enterprise users and over 5 million completed actions and queries. The company also said revenue has grown 50-fold over the past two years and that revenue bookings have exceeded $150 million.

Those bookings should not be read as annual recurring revenue. Chatterjee said the figure includes the total value of multiyear contracts, including two- and three-year agreements, and declined to provide Ema’s annualized revenue run rate. The company also claims that more than 90% of customers have expanded beyond their initial use case and that its net dollar retention is approximately 180%.

Ema said it maintains gross margins close to 80% and charges for completed tasks and business outcomes rather than software seats or AI-token consumption. Both the margin and customer-expansion figures come from the company and were not independently verified in the supplied reporting.

The other cluster sources—The AI Insider and Indian Startup News—confirm the headline funding event and Creaegis’s role, but their full article text was unavailable. The detailed operating claims in this report therefore rely primarily on TechCrunch’s interview-based account and Ema’s own disclosures.

Why the round matters for enterprise AI

Ema’s funding reflects a shift in the enterprise AI competition. Vendors are no longer competing only to add a copilot inside an existing application; they are increasingly trying to control the workflow spanning several applications. That distinction matters to product teams because the value of an agent depends on whether it can safely complete a process, not merely produce a useful response.

The approach also puts pressure on the traditional boundaries between software and services. If an AI system can interpret a request, retrieve data, apply policy, update multiple systems, and leave an auditable record, some work currently sold as configuration, integration, or routine operations could be packaged as an outcome-based service.

However, replacing enterprise software is a much higher bar than connecting to it. Businesses still need systems of record, permissions, compliance controls, monitoring, and reliable recovery when an agent encounters an exception. Ema’s own strategy—initially wrapping around existing applications—acknowledges that customers are likely to adopt automation incrementally rather than discard core systems immediately.

Ema also operates in a market where larger AI companies are moving closer to enterprise operations. TechCrunch noted Anthropic’s expansion into financial and legal workflows and OpenAI’s use of forward-deployed engineers to put AI into production. Chatterjee does not regard frontier labs as direct competitors, arguing that improvements in their models benefit Ema’s orchestration layer. That view will be tested as model providers add more integrations, workflow tools, and enterprise support themselves.

What to watch next

The clearest signal will be whether Ema converts its reported bookings into durable, recurring revenue and whether customers continue expanding deployments after pilot programs. Buyers should also look for evidence of task completion quality, exception handling, auditability, and the human review required for sensitive HR, finance, and IT actions.

Ema’s international expansion will provide another test. Entering Asia-Pacific, South America, and the Middle East will require local compliance knowledge, language support, implementation capacity, and integrations with regional systems—not just additional sales coverage.

The market should also watch how Ema prices against traditional SaaS licenses and consulting engagements. Outcome-based pricing may align the vendor with customer results, but it can make costs harder to forecast if workflow volumes change. Gross margins will likewise depend on model costs, integration maintenance, support, and the amount of human intervention needed when automation fails.

Creati.ai perspective

Ema’s round is significant less because it proves that enterprise software is about to disappear than because it shows where investors and buyers see a possible replacement layer forming: between employees and the applications they use. The company’s strongest argument is its focus on multi-step work rather than single-turn assistance.

The harder question is operational. Enterprise customers will pay for agents that complete important processes reliably, explain what they did, and fail safely. Ema’s next phase will show whether its reported expansion and margin profile can hold as deployments become broader, more regulated, and more dependent on production-grade controls.