Using Generative AI to Empower the Enterprise

Saket Srivastava doesn’t frame AI as buzz or hype. He talks about it as a strategic engine reshaping how teams work, how organizations plan, and how technology leaders connect data to outcomes. For the CIO of Asana, a global collaborative work management platform used by more than 130,000 customers in over 200 countries, generative AI is not merely a productivity add-on; it’s a tool that amplifies human decision-making and unlocks new organizational capabilities.

Srivastava explored the current hype vs. reality of AI, how Asana’s data foundations uniquely position it to leverage generative models, and what use cases are already creating tangible value, both inside the company and in the product itself.

Separating Hype from Long-Term Potential

Generative AI has been the topic of countless boardroom conversations and technology roadmaps, but Srivastava urges leaders to take a balanced view. While the technology may feel overhyped in the short term, he believes the long-term possibilities are underhyped because organizations are still just beginning to understand what generative AI can truly do.

“Gen AI might be overhyped in the shorter term, but in the longer term, it might actually be underhyped. The possibilities with Gen AI are something we are just beginning to understand. We're still scratching the surface, I think this is still the preview that we're seeing.”

This perspective influenced how Asana approaches AI, with curiosity and ambition, but grounded in realistic expectations and strategic use cases.

AI Built on a Foundation of Work Data

A core strength Asana brings to generative AI stems from its existing data model: what Srivastava calls the work graph. Because Asana retains the atomic unit of work within its platform, the quality and fidelity of that data make it particularly well-suited for AI augmentation.

Rather than generic text data or disconnected spreadsheets, Asana’s work graph encapsulates structured information about who is doing what, when, how tasks interrelate, and how progress toward organizational goals is tracked. Srivastava notes that this kind of data enables stronger outcomes from AI models: with high-quality inputs, generative insights become more actionable and contextual.

That provides fertile ground for applications that go beyond personal productivity into organizational productivity.

From Personal Copilots to Organizational Planning

Srivastava frames generative AI use cases at Asana in two broad categories:

Copilots for Individual Productivity

Generative AI can assist individuals within Asana by summarizing work, answering questions, and providing smart answers based on the tasks and work items they manage. Users can ask Asana questions about their workloads or progress, and the AI can synthesize responses using the underlying work graph.

This aligns with the broader trend of copilots in enterprise systems: tools that help individuals get answers and perform tasks faster without deep technical skill.

Organizational Intelligence and Planning

More transformative, Srivastava explains, is the way AI is being used to recommend organizational goals and plans. Because Asana knows the work being done across teams, it can leverage generative AI to suggest SMART goals, assist with resource planning, and even offer insights into how teams might best allocate talent based on ongoing work patterns.

“We can recommend what your SMART goals are for the company… We can help you resource plan better… we can take a look at skill sets and capabilities and make recommendations around that for you.”

This elevates AI from a tool that simply completes tasks to one that guides strategic thinking and planning based on real work data.

Internal Enablement and Business Use Cases

Srivastava also discussed how Asana is using generative AI internally to empower employees and enhance productivity across functions:

• Sales and Go-to-Market Copilots: Asana is building generative AI capabilities that help sales teams research prospects, understand customer usage, and even draft tailored messaging, effectively creating a sales cheat sheet that can speed up preparation and engagement.
• Support Deflection and Automation: Generative AI can automate answers to common support questions, reducing the load on human agents and allowing them to focus on complex inquiries.
• Knowledge Search and Answering: As companies grow, institutional knowledge often becomes fragmented and hard to find. AI helps by surfacing answers to internal questions without requiring employees to know where knowledge resides, saving time and preserving deep work.
  

Srivastava emphasizes that these AI tools are in service of humans, with accountability still resting with employees to make decisions based on AI suggestions.

“AI can be that sort of muscle and the human can be that brain and that heart… That’s sort of the way I think about this.” This balanced view, AI assisting rather than replacing, underscores Asana’s human-centered approach.

Build vs. Buy

When asked about strategy for build vs. buy, Srivastava makes it clear that there is no single answer. Instead, he advocates a use-case by use-case approach that accounts for company context, industry, and regulatory environment.

Asana does not build its own foundational models; rather, it leverages existing models from partners such as OpenAI and Anthropic — ensuring access to innovation without reinventing core AI infrastructure. However, the company does choose to build specific use cases (such as the sales copilot) tailored to its business needs.

Srivastava’s philosophy reflects a hybrid model: stay close to vendor innovation while also experimenting and creating custom implementations where strategic value is clear.

Preparing for AI’s Next Wave

Looking ahead, Srivastava suggests that the next phase of generative AI will move beyond pull-based interactions, where users request assistance, toward proactive, context-aware recommendations that anticipate needs.

This shift from reactive to proactive AI represents a fundamental change in how organizations interact with technology: from querying AI for help to having AI signal opportunities and risks as they arise.

Lessons Learned

Across his conversation, Saket Srivastava shared several strategic lessons for enterprise technology leaders embracing generative AI:

• Manage expectations: Short-term hype should not distract from thoughtful, long-term value creation.
• Build on strong data foundations: High-quality, structured data (like Asana’s work graph) amplifies AI effectiveness.
• Design AI for humans, not to replace them: AI should augment human decision-making, not displace accountability.
• Tailor build vs. buy decisions: Use cases should drive decisions about what to build in-house and what to partner on.
• Think proactively: Future AI will go beyond responses to become anticipatory, offering insights before users ask.
  

In a world where enterprises grapple with both AI’s promise and complexity, Srivastava’s approach offers a roadmap that blends pragmatism, innovation, and strategic ambition, all grounded in real business needs and human outcomes.