The news
Ema, a startup that builds what it calls “AI employees,” said on September 23 that it raised a $77 million Series B. Creaegis, a Bengaluru-based investment firm, led the round, and existing investors Accel, S32 and Prosus increased their stakes. Ema said total funding is now $140 million and that its valuation more than quadrupled from its prior round; it did not give a figure.
The product coordinates multiple AI agents to handle multi-step internal processes, such as employee questions, IT tickets and finance tasks, and works across existing applications, which Chatterjee told TechCrunch some customers later scale back or replace. Ema named Wipro, Hitachi, ADP and PwC as customers, and TechCrunch reported the list also includes Google, Microsoft, KPMG and NTT DATA.
The company’s metrics are its own. It said revenue grew 50-fold over 24 months and that one global systems integrator uses the platform to field about 2.9 million employee queries a year across 240,000 staff. TechCrunch reported more than 50 active enterprise deals, more than $150 million in multiyear bookings, net dollar retention near 180% and gross margins around 80%.
Founded in 2023 by Surojit Chatterjee, a former Google and Coinbase executive, and Souvik Sen, formerly of Okta, Ema has nearly 200 employees, according to TechCrunch. It plans to use the money mainly for sales and marketing and to expand into Asia-Pacific, South America and the Middle East.
The numbers
- Series B
- $77M
- Lead investor
- Creaegis
- Total raised
- $140M
- Multiyear bookings (TechCrunch)
- More than $150M
Why CEOs should care
Ema’s showcase deployments are telling: its release highlights a top-five systems integrator and a top-five outsourcer, and Chatterjee told TechCrunch many services companies work with Ema. That hints agentic AI may first displace the kind of back-office work companies currently pay service providers or shared-service centers to do. CIOs and CFOs should expect vendors, including their existing outsourcers, to re-price contracts around automation.
The company’s own summary is blunt, with its CEO saying enterprises “need work to get done,” not more software. For buyers, the practical questions are accuracy, escalation rates, auditability and who is liable when an agent gets it wrong. Ask for customer references at your scale before signing multiyear deals.
The bigger picture
TechCrunch noted that Anthropic and OpenAI are pushing deeper into enterprise work, but Chatterjee told the outlet he does not see frontier labs as direct competitors, since Ema draws on more than 150 models and focuses on integration and orchestration.
What “Fact-checked” means
Fact-checking means testing a story’s facts against the evidence before it is published. This story went through at least two separate checks before this version was published.
- What we checked
- Its names, figures, dates, job titles, quotes and who said what were checked against the story’s sources, including its main source where it could be opened. The headline was checked for accuracy and overstatement.
- How
- A first check reviewed the whole story. If it passed, a second, skeptical check went back to the sources to look for mistakes in the most important facts. If a check flagged the story, it was edited to fix the problems found, and a separate re-check then reviewed the whole story again.
- Who
- The checks are made by our newsroom, as steps kept separate from the writing, under rules set by our editor, Hussein Mukhtar. A story the checks still flag is held for the editor, who decides whether it is fixed, published or dropped.
- If something is wrong
- “Fact-checked” does not mean error-free. If a material error is found after publication, we correct the story and add a note saying what changed. Report an error







