Skip to content
2 min read

Beltic x Keiki: so merchants can trust the agents at their door

We're partnering with Keiki to put verified agent credentials to work in real commerce flows, so merchants can tell a trusted agent from a bad actor.

Luca Castellanocompanypartnership

We're partnering with Keiki so merchants can trust the agents at their door.

Agents are already shopping, booking and buying on our behalf. The open question is trust.

Merchants can't tell a good agent from a bad bot

Every day, more of the traffic hitting a merchant's storefront is not a person behind a screen. It's an agent acting for one.

Merchants spent years building fraud filters to keep bad bots out, and those filters work. The problem is they can't tell a bad bot from a good agent, so they block both. Real customers get declined, real orders get dropped, and a whole new channel of commerce gets treated like an attack.

Credentials only matter if agents actually carry them into real transactions. That's why we're partnering with Keiki.

Why Keiki

Keiki builds agents people actually use. Their assistant lives in iMessage and handles real work for its users, and their platform lets any team launch customer-facing agents across iMessage, WhatsApp, Slack, Telegram, SMS and email. These are agents that answer questions, complete workflows and take action for real people and businesses, inside the conversations where commerce already happens. Keiki is backed by Y Combinator and, like us, they build with a bias for shipping.

Beltic is the identity and verification layer for AI agents. Know Your Agent. We issue verifiable credentials that let a merchant check, in real time, which agent is at the door, who it acts for and whether it should be trusted.

What we're building together

Through this partnership, we're putting verified agent credentials to work in real commerce flows. We're starting with a joint proof of concept: agents built on Keiki's stack carrying Beltic credentials, so a merchant can tell a trusted agent from a bad actor in milliseconds, and good agents never get caught in the fraud filters.

The world we're working toward is simple. Legitimate agents should be recognized and accepted by merchants, not blocked. When that happens, merchants open a new revenue channel without opening themselves to fraud, and people can finally delegate the shopping, booking and buying they already want to hand off.

Where this fits

We came out of stealth in late September, backed by $8.8M from Norwest, Collide Capital and other investors, with a team from Stripe, Coinbase and other fintech and infrastructure companies. This partnership is what that foundation is for: working with teams who ship agents in the real world, so trust becomes part of the transaction itself.

If you're a merchant looking to safely accept agents, come talk to us.