Databricks co-founder outlines playbook for landing large enterprise deals

Startups today face intense pressure to scale quickly, and the fastest path often runs through large enterprises with deep IT budgets. At TechCrunch Disrupt, Arsalan Tavakoli-Shiraji, co-founder and senior vice president of field engineering at Databricks, shared how the company moved from a fledgling data-analytics startup to an AI and big-data giant valued at $190 billion with more than 70 percent of the Fortune 500 as customers. His message to founders was clear: winning those contracts requires more than a superior product.

Tavakoli-Shiraji warned that many early-stage founders assume the technology will sell itself. He recounted a meeting with a prospect who had already signed a multi-year deal with a competitor and was unwilling to listen. He also described early negotiations that stalled because the startup was unprepared for the legal, contractual and security questions typical of enterprise buyers. Those experiences taught him that people and processes matter most when courting large organizations.

A turning point for Databricks came when the company secured a partnership with Microsoft and became an Azure cloud partner. By aligning its sales agreements with Microsoft’s terms, Databricks gained access to Microsoft’s sales force, which began booking meetings and advocating on its behalf. “All of a sudden, that meeting wasn’t hard to get,” Tavakoli-Shiraji said. The partnership opened doors to new prospects, even though it also generated requests for features in regions where Databricks had not yet expanded. He noted that today Databricks sells its platform to startups such as Hightouch, Addepar and Tecton under its own terms and cybersecurity framework, illustrating how a larger partner can become a launchpad for later independent sales.

To help founders navigate enterprise objections, Tavakoli-Shiraji grouped concerns into four categories he calls the “four Cs”: context, choice, cost and control. Context refers to AI that can be trained like a new employee, absorbing a company’s terminology and processes. Choice means giving customers the ability to switch AI models rather than being locked into a single provider. Cost focuses on predictability of operating expenses, and control covers safeguards that protect data and govern AI behavior. He also observed that most CIOs rate their AI adoption at the low end of a five-point scale, indicating that many enterprises are still in the exploratory phase and that meaningful AI deployment will take years.

Finally, Tavakoli-Shiraji highlighted the untapped opportunity in modernizing legacy systems. He explained that many banks still run on IBM AS/400 mainframes, and few vendors are willing to touch that code. AI agents, however, can analyze and decompose legacy code, enabling new interfaces or complete replacements. For startups, this creates a pathway to deliver high-value solutions that address costly, entrenched technology stacks.

Founders looking to break into the enterprise market should therefore focus on building strong relationships, aligning with established partners, and addressing the four Cs in their product narratives. While the appetite for AI is growing, the timeline for widespread, value-creating adoption remains long, suggesting that patient, process-driven go-to-market strategies will be the most rewarding.