Why We Built Reelin ID: The Case for Human-Anchored Agent Identity

Swiftdroom helps people use AI to compete more effectively in a world where job hunting is increasingly a volume game. The premise is simple: let AI handle the repetitive work of tailoring and submitting applications, so candidates can focus on the conversations that actually matter.

What we quickly discovered is that as AI agent volume scales, the platforms on the receiving end face a trust problem they have no infrastructure to solve. Every inbound automated request, whether it comes from an authorized agent acting for a real candidate or from an anonymous bot farm, looks identical from the platform’s side.

The infrastructure gap

The fundamental question we kept returning to: how do you prove that an AI agent’s action was authorized by a specific, verified human, not just claimed, but cryptographically proven? Not just a human somewhere, but this specific human, whose identity has been verified, who explicitly authorized this exact action.

No such infrastructure existed. So we built it.

The design constraints

Users should never have to touch a wallet. The web3 onboarding flow is a conversion killer. Asking someone to install MetaMask, buy ETH, and sign transactions is a non-starter for a typical user. The identity infrastructure had to be completely invisible.

The trust had to be real. Not self-reported (“I’m a human”), not CAPTCHA-based (“I solved a puzzle”), but verifiable against a real-world identity. That meant government ID verification, the same kind that financial institutions use.

The architecture had to be decentralized enough to be trustworthy long-term.If Reelin controls the entire trust stack, then trusting a Reelin-verified agent means trusting Reelin. That’s not infrastructure, that’s a vendor dependency. We needed on-chain anchoring so the system works even if Reelin disappeared.

What we built

Reelin ID is a decentralized identity-forking engine. A human authenticates with their social identity and verifies their real-world identity with a government document. Reelin creates an Ethereum wallet for them server-side and issues a DID, a permanent, unique decentralized identifier. The user never sees the wallet. They just have an identity.

When they activate an AI agent, Reelin generates an Ed25519 keypair for that agent. The public key is stored on-chain in the ReelinRegistry smart contract. The private key goes to the agent. From that point on, every action the agent takes is signed with this key, and any platform can verify the signature against the on-chain record in under 50 milliseconds.

The broader picture

Swiftdroom is the first integration. But the pattern is universal: anywhere an AI agent acts on a human’s behalf, there is a trust problem that Reelin ID solves.

Healthcare scheduling agents. Legal document submission bots. Financial services automation. E-commerce procurement. Every one of these domains is going to face the same immune response that job boards are facing today, and they will need the same infrastructure to resolve it.

Reelin ID is that infrastructure. We are building it in public, starting with the problem we know best, with the ecosystem we have already built.

If you are working on the same problem from another angle, whether you are a platform dealing with automated traffic or a developer building agents, we’d love to talk. Reach out at hello@reelin.id.