Ema, the Bengaluru-and-San-Francisco startup founded by ex-Coinbase and Google product executive Surojit Chatterjee, closed a $77 million Series B on September 23, led by Creaegis with Accel, S32, and Prosus all increasing their stakes. The round more than quadruples the company’s valuation from its previous raise and brings total funding to $140 million.
The product itself is not a new category. Enterprises have bought workflow automation and AI-assisted service desks for years. What’s notable is what Ema refused to call it: not a SaaS platform, not a copilot, not an “AI agent” in the now-crowded sense of the term. The company’s language, from the funding announcement down to how its team describes deployments, is “AI Employees” - a framing that puts the product in competition with headcount and BPO contracts, not with other software line items.
A naming decision that changes who signs the check
“Software” gets evaluated by IT procurement against a feature checklist and a security review. “Employees,” even artificial ones, get evaluated by a business unit leader against a much simpler question: does this replace work I currently pay people or a vendor to do, and how much cheaper or faster is it. Ema’s own disclosed numbers lean into exactly that comparison - over a million IT service tickets handled annually at large BPOs, and a deployment at a $50 billion conglomerate that went from pilot to production in four weeks across more than 20 systems, cutting ticket volume by 30%.
Whether or not “AI Employee” survives as an industry-standard term, the positioning bet behind it is a legible one: naming the product after what it replaces, rather than after the technology underneath it, changes which budget it competes for and which executive owns the buying decision. A CIO evaluates software. A COO evaluates headcount. Those are different conversations, with different urgency and different price sensitivity, and Ema built its go-to-market to have the second one.
The quieter signal is the founder, not the funding
Chatterjee’s own path - IIT Kharagpur to Google to Coinbase’s product organization through its 2021 IPO to founding Ema in 2023 - is the kind of resume that gets a founder taken seriously by enterprise buyers before the product has proven itself, which matters enormously in a sales motion built on displacing incumbent vendors and headcount. It does not substitute for the deployment numbers the company is citing, but it is very likely part of why a $50 billion conglomerate let an early-stage vendor near twenty of its systems of record in the first place.
The larger pattern for founders building anything adjacent to the current AI-agent wave: category language is not a marketing afterthought layered on top of a finished product. Ema chose its comparison set - human employees and outsourced service contracts, not software subscriptions - years before this raise, and that choice appears to be doing real work in who takes the sales call and how fast they say yes.