We made AI a seamless part of daily workflows with GrabGPT

GrabGPT’s security is equally important. Running on Grab-owned infrastructure ensures sensitive company data stays protected. We can audit the platform to prevent misuse, allowing Grabbers to harness Generative AI without compromising privacy or company processes.

Additionally, GrabGPT is available to Grabbers everywhere, regardless of where they’re based.

Bottom-up innovation

Grab embraced AI tools early, encouraging Grabbers to find ways to integrate them into their workflows. GrabGPT emerged from this ethos—not as a top-down directive, but as a happy accident.

In March 2023, the machine learning platform team began exploring Large Language Models (LLMs) to reduce the burden of manually answering user requests through an LLM-powered support bot. While the initial goal wasn’t achieved immediately, our team discovered broader possibilities. We channeled the momentum to create GrabGPT, a multi-purpose Generative AI interface for the entire company.

(Also read: From failure to success: The birth of GrabGPT, Grab’s internal ChatGPT)

GrabGPT’s transformative power

GrabGPT sparked a broader conversation about the potential of LLMs at Grab. It’s a testament to how a small group of engineers, equipped with the right tools, can create something transformative.

We’ll continue expanding GrabGPT’s capabilities to ensure it stays useful across departments. Meanwhile, the machine learning team is making progress on the original support bot project, which is now on track for success.

Today, GrabGPT is more than a tool—it’s a symbol of experimentation and adaptability at Grab.