Stampli, the B2B accounts-payable platform, cut onboarding time by 68% after weaving ChatGPT Work into its workflow.

Why the internal angle matters

Most headlines tout generative AI for chatbots, code generators, or image designers. Stampli shows the same tech can speed up a service operation. Launching a client faster frees implementation teams, shortens the revenue-recognition cycle, and boosts satisfaction before the product even runs.

The context behind the numbers

Typical AP rollouts require data imports, rule setups, user-role assignments, and a bundle of client-specific documents. In many SaaS shops these steps drag on for weeks, chewing up senior staff time that could be spent on higher-value work. Stampli spotted the repetitive checklist items and fed them to ChatGPT Work, pairing internal prompts with workflow automations.

The company didn’t just flip a switch. It built a library of prompts and automations that shepherd the model through each onboarding stage. The result is a repeatable process that can be launched for every new client, not a one-off experiment.

Risks and limits

Automation still needs human oversight. Complex contracts, odd tax jurisdictions, or legacy ERP links demand expert judgment. Letting AI draft documentation without review can seed compliance errors. The system’s success also depends on well-crafted prompts; bad instructions yield misleading output and force teams to waste time fixing mistakes.

Steps other teams can take

If you lead a customer-success or implementation group, try these steps to gauge an AI boost:

  • Map every action from contract signing to go-live.
  • Log how long each step takes.
  • Highlight tasks that repeat unchanged for each client.
  • Spot where documentation or configuration is manually rewritten.

Targeting repeatable, low-complexity work is where AI shines. Start with a small pilot, refine prompts from feedback, and expand only after the model proves reliable.

What to watch next

Stampli’s rollout is still early, and the impact on churn, upsell velocity, and cost structure remains unclear. Observers will watch whether the speedup turns into measurable revenue gains or simply reshapes internal workloads. As more vendors test AI in service pipelines, the line between product innovation and operational efficiency will keep blurring.