The Fabric builds the company before the pitch.

In early 2015, an engineer in Chennai built a tool to test software network functions in the cloud. It had a sensible, unglamorous job: help other engineers check their work. Two startups, VeloCloud and Appcito, licensed it. Then the people behind The Fabric began asking whether the tool could support a business of its own.

That became the starting point for Pensa. A piece of internal infrastructure had acquired a second identity: evidence of a market. It is a useful place to begin understanding The Fabric, because the company’s work often happens before the familiar startup story begins.

First, find the missing piece.

The founding observation was straightforward. Applications were moving into the cloud. Networking, security and operations would have to follow. Each transition would leave awkward gaps between what enterprises needed and what their existing systems could do.

Rajan Raghavan and Prabakar Sundarrajan had worked together at Ankeena, which Juniper acquired. In 2012, they joined Sumant Mandal to establish The Fabric. Their accumulated experience suggested several potential companies, each tackling a particular infrastructure problem. Building one business to handle the whole transition would have been a different undertaking.

The Appcito origin story shows the approach in miniature. In 2013, The Fabric was exploring an application-level front end for cloud software. Conversations brought in entrepreneurs Alak Deb and Kamal Anand, followed by Siva Mandalam, whose experience included security at F5 Networks. The proposition became more specific as the team took shape. Appcito eventually joined A10 Networks in 2016.

“What we need is mutually compatible, driven entrepreneurs who share our vision.”Rajan Raghavan · April 2018

The work before the check clears.

The Fabric occupies the stretch between an interesting technical possibility and a company ready for institutional venture funding. Its founders help choose the problem, find people equipped to solve it, test the proposition and assemble the business. A founder arriving with domain knowledge can gain collaborators who have already dealt with product development, recruiting and enterprise buyers.

VeloCloud began with a question about applying cloud networking to the enterprise wide-area network. The Fabric worked with Sanjay Uppal, Steve Woo and Ajit Mayya. Its 2014 account describes discussions with prospects, potential partners and industry figures to refine the opportunity. The announcement also named $21 million in venture funding from The Fabric, NEA and Venrock. That amount belonged to VeloCloud’s financing, rather than The Fabric’s own budget.

Engineering is part of that process. The Fabric’s Innovation Lab in Chennai describes participating from market research through prototyping and development, with field feedback influencing feature priorities. Pensa makes the arrangement tangible: a tool built to help two companies suggested an opportunity for another. Shared technical work could produce something besides a cheaper prototype.

Three million dollars is a beginning.

In a May 2018 interview, Raghavan described formation capital followed by seed investment of up to $3 million. He put the journey to a marketable product at about nine months, followed by initial trials and another fundraising effort. Those were historical descriptions of the model, with all the execution work still attached.

The distinction matters when asking what this cost. Seed money funds a particular company. The Fabric also raises pools to support several companies: Fabric 3 closed at $20 million in April 2018, with backers including Verizon Ventures, March Capital Partners, Citrix and Hitachi. The economic logic is venture investment: commit money and company-building effort in pursuit of valuable ownership stakes. Founders evaluating such an arrangement need to negotiate ownership and responsibilities alongside the check.

The other end of the calculation appears in VMware’s financial reporting. Its purchase price for VeloCloud was $449 million, net of cash acquired. That was the transaction price for VeloCloud; it does not reveal The Fabric’s proceeds. Ownership, dilution and investment terms determine what any individual backer receives.

The customers behind the founders.

The Fabric’s immediate collaborators are entrepreneurs. The products they build serve a second audience: enterprise technology teams wrestling with unreliable connections, application delivery, security and operational complexity. Investors form a third constituency. Keeping those groups separate helps explain the business. An enterprise purchases a portfolio company’s software; a founder works with the foundry to create that company.

OpsCruise illustrates the downstream need. Co-created in 2018, it used telemetry and machine learning to help teams understand cloud application behavior and investigate performance problems. When Virtana announced its acquisition in June 2023, it named Avis, Telefonica, Tesla and United Airlines among OpsCruise users. Those were customers of the portfolio business, rather than a client list for The Fabric.

The distinction from a conventional investor lies in how early and deeply The Fabric participates. The useful comparison for an entrepreneur is practical: what expertise will actually join the project, who will build the first product, and who can introduce credible buyers? A studio’s name does little work by itself.

Borrow the questions.

The portable lesson is the sequence. Find a costly problem created by a technical shift. Discuss it with people who might buy the solution. Recruit around the knowledge missing from the room. Build something that makes the conversation concrete. Let the response change the product.

That sequence depends on access and fit. Infrastructure veterans can ask informed questions and reach relevant buyers. A team without those relationships must earn them. A founder who wants complete independence may find an involved co-creator uncomfortable. A consumer business may need expertise quite different from cloud networking. Shared resources help only when they address the actual bottleneck.

The products can outlast the original partnership. In July 2025, Arista acquired VeloCloud’s SD-WAN portfolio from Broadcom. Years after the first question about cloud networking, the answer was still changing hands. Pensa began with a smaller clue: two teams already found a tool useful. Before asking how large a company might become, someone noticed work that needed doing.