Ema raises $77M to replace enterprise SaaS with AI agents
Original: Ema raises $77M as AI starts eating into enterprise software and services
Why This Matters
AI agent startups are now directly targeting SaaS budgets—a multi-hundred-billion-dollar market that incumbents like Salesforce and ServiceNow have long dominated.
AI agent startup Ema closed a $77M Series B led by Creaegis, bringing total funding to $140M. Founded in 2023 by ex-Google and Okta executives, Ema deploys multi-agent 'AI employees' to automate HR, IT, and finance workflows—and aims to displace traditional SaaS products entirely.
Ema, a 2023-founded startup that automates enterprise workflows using coordinated teams of AI agents, has raised $77 million in a Series B round led by Bengaluru-based Creaegis. Existing investors Accel, Section 32, and Prosus all increased their stakes. Total funding now stands at $140 million, and the company says the new round more than quadruples its valuation from its 2024 raise—though it declined to share the exact figure. The round was pure primary equity, with no debt or secondary sales.
Co-founded by former Google and Coinbase executive Surojit Chatterjee and ex-Okta executive Souvik Sen, Ema positions its technology as 'AI employees'—systems that orchestrate multiple agents to execute multi-step business processes across a company's existing app stack. Chatterjee's stated ambition goes further than automation: he expects customers to eventually drop legacy SaaS products altogether. 'Many of our customers are already on the way to replace large SaaS applications completely,' he said, arguing those platforms are increasingly functioning as little more than databases.
Ema's strategy is to first wrap around a company's existing tools, then progressively reduce dependence on them. The platform can draw on over 150 models—frontier and open-source—while Ema handles the domain knowledge, integrations, and orchestration layer. Chatterjee views Anthropic and OpenAI not as competitors but as model suppliers: 'Progress in frontier models is actually very beneficial to us.'