As AI continues to reshape the software industry, Asana’s journey toward integrating AI into its core offering demonstrates how a company can adapt quickly to emerging technologies while maintaining its strategic vision. In a recent Breaking Changes episode, Cliff Chang, AI Engineering Lead at Asana, shares the process of Asana’s AI transformation, starting from exploratory projects to becoming a core component of the company’s mission.
This narrative offers a unique look into how Asana embraced AI through rapid prototyping, organizational flexibility, and strategic leadership—all while navigating the uncertainties that come with integrating such a transformative technology.
Here’s how Asana adapted to the AI landscape, offering valuable lessons for leaders aiming to do the same.
From CEO mandate to team execution
Asana’s AI journey began with decisive leadership. The company’s CEO, witnessing the rapid advancements of models like ChatGPT-3 and GPT-4, recognized that AI’s impact was accelerating beyond expectations. He saw the potential for AI to redefine Asana’s offerings and took immediate action, instructing the head of engineering and product teams to establish an AI organization, not in the distant future, but “now.”
Interestingly, Chang wasn’t part of this AI initiative at first. He was originally slated for a significant promotion to lead a non-AI division, a role more aligned with his established strengths. However, he felt drawn to the AI project, recognizing it as an opportunity to be closer to the cutting-edge work he was passionate about. By requesting to be part of the AI organization instead, he shifted his career trajectory in a way that ultimately proved to be pivotal for Asana’s AI journey.
Advice for leadership: When exploring disruptive technologies, empower your team members to pursue projects that align with their passions. This often leads to greater innovation and commitment.
Rapidly prototyping AI’s potential
Rather than jumping headfirst into large-scale AI projects, Asana took a pragmatic approach. Within two months of ChatGPT’s release, the AI team began prototyping various AI features. Starting with a small, focused group of four engineers and a product manager, they explored over 20-30 potential applications of AI within Asana’s platform. This rapid prototyping phase allowed the team to quickly test and evaluate what AI could bring to their product.
According to Chang, these prototypes were developed over two to three months, and the process wasn’t about building polished features but rather about quickly understanding what was possible. The team’s ability to try out these concepts rapidly within Asana’s internal environment demonstrated the potential of AI to enhance the platform’s functionality.
Advice for leadership: Encourage rapid prototyping and experimentation when working with emerging technologies. Quick feedback loops can help identify high-potential applications and areas of improvement early on.
Establishing a centralized AI team
Instead of distributing AI responsibilities across different departments, Asana chose to form a dedicated AI organization. This decision wasn’t made lightly. Chang explained that the structure wasn’t intended to be permanent; it was an intentional, strategic choice to accelerate the development of AI expertise. The goal was to create a concentrated hub of talent, which could later disseminate AI knowledge and skills across the broader organization.
“If we spin up an AI organization, this doesn’t have to be a forever thing… It’s more efficient to get to a place where every engineering team is good at building with AI by having an AI organization for a while and then redistributing.”
—Cliff Chang
By centralizing AI efforts, Asana could focus on building deep expertise without being hampered by varying levels of interest or skill across different teams.
Overcoming skepticism and building alignment
Chang acknowledged that not everyone at Asana was immediately convinced of AI’s potential. The company’s advantage was a CEO who understood the complexities of AI, having invested in AI-focused companies. This top-down understanding provided clear strategic direction and ensured that AI was a core part of Asana’s vision, instead of an experimental side project.
By concentrating AI efforts within a dedicated team, Asana allowed early adopters and enthusiasts to drive progress. This approach gave skeptics the opportunity to see tangible results over time, gradually building organizational alignment.
Advice for leadership: Both top-down vision and grassroots enthusiasm are essential when navigating disruptive technologies. Be willing to champion AI while giving teams the autonomy to explore and demonstrate its value.
Developing AI talent
In a market where AI talent is scarce and highly competitive, Asana opted to develop expertise internally. Instead of hiring external AI experts, they focused on leveraging the existing strengths of their engineers, who already possessed deep product knowledge. Given that Asana was not building its own AI models but leveraging pre-existing ones from companies like OpenAI, this approach allowed for practical, user-focused AI solutions.
Chang noted that many so-called AI experts only have a few months of experience, given the field’s rapid evolution. Therefore, nurturing internal talent provided Asana with a more grounded, product-oriented perspective on AI applications.
Advice for leadership: Invest in developing AI skills within your team. Building this expertise from within ensures that your team’s knowledge is deeply rooted in your product and customer needs, leading to more practical and effective AI solutions.
Preparing for agentic AI
Chang’s “hot take” on AI’s future impact centered around the idea of “agentic AI”—AI systems capable of autonomously executing complex, multi-step tasks. He anticipates that this will fundamentally change how businesses operate, enabling rapid scaling and altering traditional company dynamics.
“Agentic AI means you can scale up and down your team extremely quickly… I still struggle to wrap my head around what would happen if someone can just hire 1,000 agents to do something.”
—Cliff Chang
Preparing for agentic AI requires not just technological readiness but also a willingness to adapt business models and operations to leverage AI’s full potential.
Lessons from Asana’s AI journey
Asana’s experience integrating AI offers valuable lessons for organizations at any stage of their AI journey. Their success lies in a balanced approach that combines visionary leadership, rapid experimentation, and adaptable structures.
Key lessons for tech leadership include encouraging passion-driven paths, embracing rapid experimentation, developing internal AI expertise, maintaining organizational flexibility, and championing a clear vision. Asana’s journey provides a roadmap for navigating AI adoption, from initial exploration to full integration, offering insights that are relevant for any organization aiming to harness the transformative power of AI.
For more of Cliff Chang’s insights, be sure to check out the full episode, “Crafting Business Strategies for Seamless AI Integration.” Learn more wisdom from industry experts by subscribing to Breaking Changes on Apple, Spotify, and YouTube.