Strategies for building AI tools people will actually use at work

Jackie Rocca is VP of Product, AI, at Slack, where she oversees the vision and execution of Slack AI, which brings generative AI natively and securely into Slack’s user experience.

The application of generative AI in the workplace is a long game, and we’re in the very early stages. Most businesses looking for ways to leverage AI are in uncharted territory — either experimenting with pilot programs or still exploring ways to meaningfully incorporate the technology into their daily operations. That means the challenge is on for product leaders who are building (or thinking about building) AI tools that employees will actually love and use, whether that’s an app integration, a chatbot, or an AI experience built natively into your existing product offering.

I’ve held numerous product roles throughout my career, but my current work leading the development of Slack’s AI capabilities is my most exciting one yet. Our product teams have kept boots on the ground, listening to what customers care about and thinking hard about how AI can help them meaningfully reach their productivity goals. That’s led us to develop a set of concepts that I believe will help other product teams stay grounded amid all the AI hype and buzz.

Here is a guide that may be helpful for leaders as we build these new products and incorporate large language models into our business strategies and tooling.

Root your gen AI application in users’ needs

An effective way to encourage the widespread adoption of AI tools is to integrate them seamlessly into employees’ existing flow of work.

When building workplace AI tools that resonate with people, the starting point should always be the users themselves. Rather than adopting a broad technology-first approach and asking, “What can we do with AI?” work backward to home in on your core user problems first.

For example, the issues that occur in people’s working lives can vary, but I commonly see complaints of information overload, the inability to optimize or effectively utilize your knowledge corpus, and getting bogged down with mundane to-do lists and tasks that just feel like a waste of time. How can you apply generative AI to solve these user problems and create a more efficient, human, and enjoyable work environment?