Early on, Databricks co-founder and SVP of field engineering, Arsalan Tavakoli-Shiraji, learned that having great technology alone doesn’t land enterprise deals. At TechCrunch Disrupt on October 26, 2026, he shared how younger companies often stumble by overlooking people, processes, and enterprise-grade preparedness while believing their product will sell itself. Databricks itself—now a giant in AI and data, worth around $190 billion and serving over 70% of the Fortune 500—started small, with those early lessons shaping its growth strategy.
Why selling through partners can open doors
Tavakoli-Shiraji recalled how his team failed to get a foot in the door with a major Swiss bank until partnering with Microsoft as an Azure cloud vendor. Under Microsoft’s umbrella, Databricks leveraged its partner network—Microsoft’s terms, legal framework, and sales force—to suddenly gain access to prospects that had been out of reach. That partnership approach helped scale Databricks’ reach while handling negotiations, contracts, and regulatory concerns around security—issues many startups aren’t ready for.
The four C’s that matter to enterprises
Enterprises don’t buy just products—they buy trust, flexibility, and clarity. Tavakoli-Shiraji groups their concerns into what he calls the “four C’s”:
- Context: AI offerings must adapt like new hires—understanding the company’s language, workflows, and priorities.
- Choice: Buyers expect to avoid lock-in; being able to switch models or vendors matters.
- Cost: Transparency around operating expenses and predictable pricing is essential.
- Control: Security, governance, and ownership of data must be clearly managed.
Tackling legacy systems with AI
Most large organizations remain stuck in “bucket one or two” of AI maturity: early experiments or basic usage, not full deployments. Big opportunity lies in modernizing legacy infrastructure—think decades-old mainframes and opaque systems no one really understands. AI can now help deconstruct these systems, expose what they do, and enable updates—from modern user interfaces to full replacements—offering a way forward for companies hesitant to touch legacy tech.
Tavakoli-Shiraji also shared stories from Databricks’ early days: the missed chances, the deal-killer questions around legal and security, and even one time when he pretended to be a lawyer to close a contract. These anecdotes underscored the non-tech challenges that can derail startup sales.
What this means
Startups hoping to win enterprise business must do more than build powerful tools: they have to align with enterprise expectations on governance, flexibility, and cost. Growth strategies that include strategic partnerships can act as force multipliers. The four C’s provide a checklist for readiness. And those that can address legacy system inertia—using AI to peel back the complexity—have a strong value proposition. For companies chasing big deals, the technical work is just table stakes; mastery of process, trust, and enterprise norms is what seals the deal.
