Free during open beta

Your AI agents should answer to you.

Ekam gives each agent you run its own identity, tied to your name. You decide what it may touch, you can read everything it did, and you can switch it off in one tap. Sign in with Google and your workspace is ready.

No invite, no card, nothing to install. Your agents can live on your laptop, in a sandbox, or on someone else's platform. Ekam holds only the identity and the off switch.

What you get

A name for every agent

Each agent you run gets its own identity and its own key. No shared passwords, and no mystery about which one did what.

Small permissions that expire

An agent that reads your calendar cannot send your mail. Each pass is for one place, lasts a few minutes, and says who gave it.

An off switch that works

Switch an agent off and every pass it holds dies within seconds. The services it was talking to are told straight away.

How it works

1

Sign in with Google

Your workspace is created on first sign-in. It is yours, not your employer's, and it is free.

2

Name an agent

Say what it may touch, in plain words. Ekam hands it a signed pass that says exactly that and nothing more.

3

Let it work, read the log

Every action is written to a log you can read. So can your agent. Switch it off any time, from anywhere.

Already have agents running somewhere? They keep running where they are. Ekam adds the identity, the passes and the off switch around them. The docs show the three calls.

Connect the agents you already run

Muse, Instinct, Wajo, Dots, the one you wrote last weekend. Each lives on its own platform, and none of them can prove to the others who it is or who it works for. Ekam gives every one of them a pass with your name on it, so they can work together and you stay in charge of what passes between them.

Musepass · owner you
Instinctpass · owner you
Wajopass · owner you
Dotspass · owner you
Your agentpass · owner you
any transport: email, chat, HTTP · any sandbox
1

Give each agent a pass

Name it and say who owns it. Ekam signs the pass; the agent keeps its own key. No shared password travels anywhere.

2

Say who may talk to whom, about what

A grant names both agents, the topics, what either may ask the other to do, and when it expires. Never a blanket "these two may talk".

3

Keep the receipts

Every exchange leaves a signed receipt either side can show and anyone can check. Switch an agent off and its passes die in seconds, on every platform at once.

Agents owned by someone else can find yours only if you choose to be findable, and you can take that back at any moment. The trust guide shows the handshake and the receipt.

Receipts, not promises

Every pass, every approval and every switch-off is written down as it happens. You can read your own log at any time, and so can an agent you allow. When two agents exchange work, each side can keep a signed receipt that anyone can check later without taking anybody's word for it.

See your usage after you sign in. How receipts work.

Your data

Built for India's DPDP law and for GDPR from the first day. The log records what happened, never the content of your messages. Everything that is yours can be exported or deleted, agents included. Hosted on Krutrim Cloud in Hyderabad with a second site in Bengaluru.

Privacy · DPDP · Security

Running agents for a company?

Bring your people in from the sign-in you already use (Google Workspace, Okta, Microsoft Entra, SAML, GitHub), give every agent in the company a named owner, and keep one audit trail for all of it. The enterprise page has the details, the standards, and a short form for a dedicated setup.

Ekam for enterprise