AI agent startup Ema raises $77M Series B

By Splitfeed EditorialLast updated Sep 26, 2026, 19:17 UTC

On September 23, 2026, Ema closed a $77 million Series B round led by Creaegis, bringing its total funding to $140 million. The company sells AI agents it describes as "AI employees" to enterprise HR, IT, and finance departments, positioning them as replacements for both legacy enterprise software and human services workflows. Ema counts more than 50 enterprise customers, among them Google and Microsoft.

The raise matters because it lands inside a broader argument about whether AI agents are generating genuine enterprise value or simply absorbing investment capital in search of a business model. Ema's framing is pointed: it does not pitch its agents as productivity tools layered on top of existing software but as outright replacements for it, which puts the company directly in the path of disputes about automation, workforce displacement, and the long-term unit economics of AI products. The presence of Google and Microsoft on the customer list sharpens that argument in both directions.

The pro-AI camp has not yet published reactions to this story. In general, supporters of AI enterprise adoption would be expected to treat the round as evidence of a maturing market, where meaningful contracts with large customers validate the product rather than just the pitch. The middle ground reaction, represented by the account Deep Thought, focuses almost entirely on that customer list: the argument is that organizations like Google and Microsoft do not name vendors for reputational reasons alone, making their presence a substantive signal of real deployment rather than marketing. Skeptics read the same facts differently. Accounts including Mike St. Clair and Sasquatch argue that the leading AI companies, including Google and Microsoft themselves, are still losing money on AI products, and that continued investment in the space without profitability points toward a correction. Mira, vixenzilla expressed a harder version of the same view, anticipating the financial failure of those same anchor customers.

The argument is unlikely to resolve without either evidence of Ema's revenue and margins or a broader shift in the enterprise AI market's profitability picture. Customer renewals, expansion beyond the current 50 accounts, and any public disclosure of unit economics from Ema or its investors would be the most direct things to watch.