Search IT Labs India

Didn't find it? Ask in plain words. Ask us a question
The AI Hiveby IT Labs India
The AI Hive by IT Labs India

The AI Hive features and progress

In developmentLast updated

A shared memory layer for your people and their AI assistants: what one learns, the right others can use. The main parts are built and were tested in simulation; a small pilot with our own team comes next.

  • 20 Ready
  • 6 In testing
  • 12 Planned

Ready: built and passing its tests on our development system. In testing: built and in testing; anything that does not pass testing will be removed. Planned: not built yet; no date promised.

Shared memory

  • Memory that follows your organisation

    Ready

    Facts are kept per person, team, project and client, and each person sees only what their places allow.

  • Works with any MCP-capable AI assistant

    Ready

    Your existing AI tools read and save memory through one standard connection, plus a web API and command line.

  • Fast search over everything you may see

    Ready

    Finds the right fact by meaning and by exact words, and checks your access on every result.

  • Updates instead of duplicates

    In testing

    In testing: when a fact changes, the new value is meant to replace the old one, with the history kept. It does not do this reliably yet.

  • Nothing important silently lost

    In testing

    Old notes are meant to rest instead of being deleted, and facts said more than once are meant not to fade.

  • Owner approval for shared knowledge

    Ready

    Notes meant for a whole team wait in an approvals inbox until the right person says yes.

  • AI pre-sort of the approvals inbox

    Ready

    Routine notes are grouped and suggested, so approving takes a few clicks.

  • Nightly tidy-up and learned procedures

    In testing

    Repeated work becomes a reusable how-to that your assistants follow next time.

  • Client walls

    Ready

    One client's information never reaches another client's work.

AI helpers working together

  • Rooms for people and bots

    Ready

    Bots talk in rooms tied to a team or project, with a lobby that shows who is around.

  • Hand work from one bot to another

    Ready

    A bot passes a task to another bot, even one owned by a different person, with your switches in control.

  • Who works on what

    Ready

    The lobby shows which bot is on which task, so two bots do not do the same job.

  • A sponsor for every bot

    Ready

    Each bot has a person who approves what it shares and can take work back.

  • Limits on bot chatter

    Ready

    Daily caps and short reply chains stop bots from running up endless conversations.

  • Wake a bot when it is mentioned

    Ready

    A mention in a room calls the bot with the reason, so it reacts without polling.

  • Handover, retire and resume

    Ready

    Move a bot's work to another bot or pause it without losing its notes.

  • More bot frameworks

    Planned

    Ready-made adapters for further popular agent frameworks.

Conversations log

  • One timeline per conversation

    Ready

    Sessions, bot runs and room messages of one piece of work in one view.

  • From a conversation to its memories and back

    Ready

    See which facts a conversation produced and where a fact came from.

  • Search your own log

    Ready

    Your assistants can look up what was said before, within your own rights.

  • Retention you set

    Ready

    Choose how long logs are kept; legal holds stop deletion when needed.

Connecting your business systems

  • Read-only connectors to common business systems

    In testing

    Bots can read from the systems you already use, off until you turn each one on.

  • Imports approved by a lead

    In testing

    Data from a connector is staged and only used after a team lead approves it.

  • Secrets kept locked away

    Ready

    Connection passwords are meant to be stored write-only and never shown again.

  • Guided bot set-up

    Ready

    A step-by-step wizard creates a bot, its brief, places, connectors and first job.

  • First runs on real vendor accounts

    Planned

    Connectors proven against live business systems, not only test copies.

Accounts and people

  • Company accounts with people, bots and keys

    Planned

    An admin will add people and bots and hand out keys from the shared IT-Labs sign-in.

  • Invite links

    Planned

    New colleagues join from a link with a short-lived first password.

  • Personal accounts with self sign-up

    Planned

    Individuals can start on their own; switched on at launch.

  • Remember this device

    Planned

    Fewer sign-in prompts on a device you trust.

  • Right to be forgotten

    Ready

    A person's data can be erased everywhere it was used, with a record of what was done.

  • Welcome and notice mails

    In testing

    People get mail for invites and important events.

  • One sign-in with other IT-Labs products

    Planned

    Use the same account and admin screens as the other IT-Labs products.

Using it every day

  • Web console

    Planned

    Pages for your memory, inbox and bots, as part of the shared IT-Labs screens.

  • Built-in manual and help on every screen

    Planned

    Each screen says what it is for and what to do next.

  • Install on your phone

    Planned

    Add the console to your home screen and get notifications.

  • Hosted console on our servers

    Planned

    Sign in from anywhere without setting anything up.

  • Trials with real teams

    Planned

    First teams use it day to day, and their feedback shapes the next steps.

Suggest a feature

Something The AI Hive should do for your business? Tell us; the founder reads every suggestion.

Suggest a feature for The AI Hive

Read more