Every so often you meet a founder whose next act is obvious in hindsight. Mohit Aron has built two decacorns. Before that, he was one of the lead engineers behind the Google File System, the distributed storage layer that made it possible to index the world’s information. At Nutanix, he collapsed the enterprise data center into software and pioneered hyperconvergence. At Cohesity, a company we were fortunate to back, he built SpanFS, a web-scale distributed file system that unified backup, recovery, and data management across on-prem and cloud, and then layered agentic search on top of it.
There is an undeniable pattern in that career – when information is scattered across systems that were never designed to talk to each other, somebody has to build the layer underneath that makes them coherent. Mohit has done it twice. He is doing it a third time, and this time he is building SciFin and pointing it at GTM teams. We are thrilled to co-lead SciFin’s $38 million seed round with our friends at Altimeter.
The GTM stack is a graveyard of point solutions
Walk into any revenue organization and count the tools. A CRM as the system of record. A conversational intelligence tool for calls. A forecasting tool. A sales engagement platform. A prospecting tool. An enablement platform. A data enrichment vendor. Half a dozen dashboards stitched on top. Each one was bought to solve a real problem, and each one solved it, in isolation. The result is that no single system knows what is actually happening in a deal and teams spend hours and days in meetings trying to answer the most basic questions since the answers live between tools and in the seams of the organization. A customer says something on a call that changes everything, and it lives in a transcript. A competitor gets named in an email thread. A champion leaves and nobody updates the account. The CRM says one thing, the rep says another, and the forecast reflects neither.
This is a structural problem, not a feature problem. For 25 years, CRM tracked the data and humans supplied the context, the judgment, and the action. We even invented a function, RevOps, whose job is to be the connective tissue between systems that refuse to connect. That was a reasonable trade when the alternative was nothing. AI has exposed that intelligence is no longer the bottleneck, context is. Point an agent at a fragmented, stale, permission-scattered data estate and it will confidently produce an answer that is incomplete or wrong. More AI on top of a broken foundation doesn’t create more confidence. It creates more output you can’t trust. Karan’s operating experience as a sales executive at first, and now as an investor in many of these GTM tools, further informed our conviction in the magnitude of the problem that exists and the elegance with which Scifin solves it.
From systems of record to systems of reality
The industry’s answer so far has been to move from systems of record to systems of action: agents that write to your CRM, draft your follow-ups, and update your pipeline. That’s useful, but it skips a step. An agent that acts on a distorted picture of the world just makes the wrong thing happen faster. What revenue teams actually need is a system of reality: a single, current, permissioned representation of what is true about every deal, account, rep, and territory, assembled from everywhere that truth lives, whether that’s a call, a document, an email, a spreadsheet, or someone’s head. Get the reality layer right, and the record and the action both fall out of it. Get it wrong, and everything above it is theater. That is the problem SciFin is solving.
The Agentic Mesh
SciFin’s core technology is what Mohit calls an Agentic Mesh: a distributed network of connected agents that continuously assemble a live context graph across the tools, people, and AI systems a revenue organization already runs on. It is, deliberately, a file system for context. It ingests from everywhere, resolves conflicts, keeps itself current, and exposes the result to both humans and agents. It gets invoked in real time and provides answers at the speed of thought. Building this at enterprise scale is a genuinely hard engineering problem, which is precisely why we think this team is perfectly suited for the opportunity. SciFin is enterprise-first by construction: multi-tenant, snapshot-aware so you can see how a deal or a territory actually changed over time rather than only where it stands today, and with role-based access control treated as a first-class citizen rather than bolted on in year three. Mohit’s founding team comes from companies like Google, Meta, Microsoft, Amazon, Apple, Nutanix, Cohesity and Rubrik.
On top of the mesh, SciFin delivers apps and skills: forecasting, prospecting, sales enablement, engagement, conversational intelligence, and deal and account intelligence, in one seamless product rather than six contracts. Each app is a wedge, and each wedge deepens the graph. The more of the workflow that runs on SciFin, the more complete the context becomes, and the more indispensable the layer underneath gets.
Founder-product-market fit
We invest in people first, and this is as clean a case of founder-product-market fit as we have seen. Karan has partnered and worked with Mohit for over a decade and watched him as he spent 20+ years building distributed systems that unify fragmented data at scale. His ambition for Scifin is characteristically large, inspirational, and timed perfectly in the market. Karan has known Mohit for years, through boardrooms and through the ups, downs and long middle of company building, which is where you actually learn what someone is made of.
We are thrilled to lead this round and partner again with Mohit and the SciFin team on building the context layer for the modern revenue organization. Let’s go!
Related Insights
