Frequently asked questions

AirCaptain is an AI implementation and operations company. We build custom AI agents inside a company's existing systems for customer experience, operations and finance teams, then stay on to run and improve them.

AirCaptain was founded by Andy Kim, Ranvir Gujral and Ozéias Sant'ana. Andy led CX and operations at Apple, eBay, DoorDash and Midi Health. Ranvir is Managing Partner of Sleeping Bear Capital and co-founded Chute (acquired 2018). Ozéias has 10+ years running ML and AI in production, including engineering leadership at The RealReal and Kevala. Meet the team.

Neither, exactly. AirCaptain does not sell a platform license, and we do not hand over a slide deck. Our engineers build the agents in your stack, and our operators run them with your team after launch.

No. AirCaptain connects to the systems you already run, including your helpdesk, EHR, POS, ecommerce and ERP. Nothing gets ripped out.

Custom. Every AirCaptain agent is built around your workflows, your policies and the systems you already run. You are not fitting your operation into a template or paying for a platform license.

Our engineers work inside your operation, not from a support queue. They sit in with your frontline leads, learn where the work actually breaks, and build directly in your systems.

Each agent works within rules your team signs off on, with clear points where a case goes to a person. We monitor every agent in production, review edge cases with your team, and adjust as your policies change.

Three steps: a free operations assessment, a free pilot on your own systems and real cases, then a full build-and-run engagement with our engineers embedded alongside your team.

The pilot is up and running on your real workflows in 1–2 weeks and runs for four weeks. You see the ROI before you commit to anything beyond it.

Access to the relevant systems, one operational owner, and a few hours a week from the frontline leads who know where the work breaks. No internal engineering is required, though we work alongside it if you have it.

Yes. The engineers and operator who run your assessment and pilot stay with you through the build and after launch.

We stay on. We keep building and expanding agents into new workflows, monitor performance, review edge cases with your team, and update the agents as your products, policies and volumes change.