Mohit Aron has built two enterprise infrastructure companies from scratch—Nutanix, which he co-founded, and Cohesity, which he founded—and both times he ran into the same problem: the data existed, but the business still ran on gut and reconstruction. SciFin Tech Inc. which launched Sept. 1, 2026, is his attempt to solve it with $44 million in seed funding.
Altimeter and Madrona co-led the round. Foundation Capital, S32 and Zetta Ventures also participated. The $44 million makes it one of the larger seed rounds in enterprise software this year at a stage when most seed rounds top out well under $20 million.
Aron described the core problem as a "context gap"—the distance between what is actually happening inside a company and what its systems show at any given moment. Pipeline reviews, forecasting calls and deal reviews at his prior companies devolved into sessions spent reconstructing the state of the business from scattered inputs rather than acting on a shared, accurate picture. "We were extracting the news and not changing the news," Aron said.
SciFin's answer is what Aron calls "context convergence." The platform pulls information from finance systems, account records, deal data, forecasts, sales rep activity, territory assignments, customer conversations and operations into a single continuously reconciled view. The system's job is to close that gap continuously, not just at the moment a user asks a question.
The user-facing layer is an AI assistant called Pixie—named after Aron's dog. The interface resembles a ChatGPT-style chat window where revenue team members can ask questions about pipeline health, forecast accuracy or deal status and receive answers drawn from the full business context the platform maintains. Users access Pixie through voice, email, Slack and WhatsApp.
Under the hood, specialist AI agents run continuously. Every time a meeting occurs, agents update Pixie's business knowledge base. As metrics shift day to day, separate agents track the changes, rebuild reports and maintain the central view. The design is meant to keep the assistant's answers current without requiring manual data entry or scheduled syncs.
SciFin is launching with an initial focus on revenue teams. That focus is deliberate. Revenue operations already spends heavily on point tools—CRM platforms, forecasting software, conversation intelligence, revenue intelligence overlays—and still produces the exact reconstruction problem Aron described. The pitch is that converging context across those tools eliminates the manual coordination layer.
Morgan Stanley Managing Director Emmanuel Dounias offered an early endorsement of the concept. "Advisors have access to a lot of information, but it isn't always connected or available when they need it," Dounias said. "SciFin's aim to bring that context together would reduce manual work, help advisors spend more time with clients on more relevant conversations, and better provide the services their clients need."
The competitive backdrop matters. Salesforce, Microsoft and HubSpot all sell AI-assisted revenue tools native to their own CRM ecosystems—Einstein, Copilot for Sales and Breeze, respectively. The incumbent advantage is integration: they already sit inside the system of record. SciFin's bet is that the problem is cross-system by nature and therefore cannot be solved from inside any single platform. Aron has been here before: both Nutanix and Cohesity were built on the argument that the incumbent stack was the wrong architecture for the problem, and both reached meaningful enterprise scale.
Altimeter has a history of backing enterprise software companies at early stages with significant check sizes, and Madrona operates out of the Pacific Northwest with a focus on enterprise and cloud infrastructure. The presence of both, plus Foundation Capital—which has backed enterprise SaaS companies including Responsys and MobileIron—signals confidence in Aron's track record as the primary underwrite rather than proven revenue metrics from a company that launched today.
Seed-stage enterprise AI rounds at $44 million reflect the current capital environment for founder-market fit plays. A repeat founder with two successful enterprise exits commands a premium at the earliest stage, and investors are willing to price that in before product-market fit is established. The risk is the same one that faces every AI-layer company: if the underlying CRM and ERP platforms improve their own AI enough, the independent context layer loses its reason to exist. Aron's counter is that no single platform owns the full context—finance, operations, customer conversations and deal data live in too many places for any one vendor to consolidate natively.
