Ema raises $77M as AI starts eating into enterprise software and services

Ema, a startup that uses teams of AI agents to automate firm processes crossed HR, IT, and finance, has raised $77 cardinal successful a caller backing circular arsenic it aims to instrumentality connected much of the enactment traditionally handled by endeavor bundle and IT services.

The Series B circular was led by Bengaluru-based task steadfast Creaegis, with existing investors Accel, Section 32, and Prosus expanding their stakes. The financing brings the startup’s full backing to $140 cardinal and much than quadruples its valuation from its past backing circular successful 2024. (Ema declined to disclose its latest valuation.) The circular consisted wholly of superior equity, with nary indebtedness oregon secondary transactions, the startup confirmed to TechCrunch.

The backing comes arsenic AI is opening to vie for dollars that businesses person traditionally spent connected endeavor bundle and IT services. Startups, large AI labs, and established bundle companies are present warring to seizure that spending.

Founded successful 2023 by erstwhile Google and Coinbase enforcement Surojit Chatterjee and ex-Okta enforcement Souvik Sen, Ema is looking to grow its presumption successful that market. The startup deploys its technology, which it calls “AI employees” — systems that coordinate aggregate AI agents. These assistance transportation retired multi-step concern processes crossed a company’s existing applications, alternatively than handling a azygous task astatine a time.

Chatterjee sees that exemplary yet reducing companies’ reliance connected accepted bundle products, including those sold arsenic software-as-a-service (SaaS). Ema archetypal “wraps” astir an enterprise’s existing applications, helium said, earlier customers tin trim their dependence connected immoderate of those products — and, successful immoderate cases, regenerate them altogether.

“Many of our customers are already connected the mode to regenerate [large SaaS applications] completely, removing dependency connected them, due to the fact that they are mostly becoming similar a database,” Chatterjee said.

Benefitting from AI labs’ push

In caller months, large AI companies person besides pushed deeper into the endeavor marketplace wherever Ema operates. Anthropic has expanded efforts to bring Claude into companies’ halfway operations, including financial and legal work. Similarly, OpenAI has established teams of forward-deployed engineers who enactment alongside customers to enactment AI into production.

Chatterjee, however, does not spot the frontier AI labs arsenic nonstop competitors. He told TechCrunch that Ema’s bundle tin gully connected much than 150 models, including frontier and open-source models, portion the startup focuses connected the domain knowledge, integrations, and orchestration needed to automate concern processes end-to-end.

“Progress successful frontier models is really precise beneficial to us,” Chatterjee stated.

Ema’s attack is already gaining traction. The startup has much than 50 progressive endeavor deals, arsenic good arsenic implicit 1 cardinal progressive endeavor users, and has handled much than 5 cardinal actions and queries. Its customers see NTT DATA, Hitachi, ADP, PwC, Google, KPMG, Wipro, and Microsoft.

Over the past 2 years, Ema said its gross has grown 50-fold, portion gross bookings person surpassed $150 million. Chatterjee said the bookings fig includes the full worth of multiyear contracts, including two- and three-year deals, alternatively than representing yearly recurring revenue. He, however, declined to disclose the startup’s existent annualized gross tally rate.

Chatterjee told TechCrunch that much than 90% of Ema’s customers person expanded beyond their archetypal usage case, with immoderate deploying the exertion crossed dozens of workflows. The startup’s nett dollar retention complaint is astir 180%, helium said, meaning existing customers are spending substantially much with Ema implicit time.

Ema is besides looking beyond the bundle itself. AI, Chatterjee said, tin instrumentality implicit immoderate of the implementation, integration, and consulting enactment that companies person traditionally paid IT services firms to execute astir endeavor software.

“A batch of the services companies are moving with us,” Chatterjee said. “They are besides dramatically changing oregon disrupting their ain concern models due to the fact that they recognize the human-forward exemplary whitethorn not beryllium the champion exemplary going forward.”

Despite taking connected enactment traditionally handled by bundle and services providers, Chatterjee said Ema has maintained gross margins of adjacent to 80%. The startup, helium noted, requires little quality enactment arsenic its AI systems larn from deployments, helping amended margins implicit time.

Ema besides does not complaint customers based connected bundle seats oregon the fig of AI tokens they consume. Instead, Chatterjee said, its pricing is tied to the completion of tasks and concern outcomes.

Much of Ema’s caller superior volition spell toward expanding its go-to-market operations, peculiarly income and marketing, aft spending its archetypal years mostly gathering the product, Chatterjee said. The Mountain View-headquartered startup has grown to astir 200 employees and has its offices successful Bengaluru, London, and Vancouver.

Ema has truthful acold focused chiefly connected customers successful the U.S. and Europe. However, it present plans to grow into caller markets implicit the adjacent year, peculiarly crossed Asia-Pacific, South America, and parts of the Middle East.