Olvid Bots

Olvid Bots are part of Olvid Enterprise and let you automate tasks within Olvid: processing incoming messages, injecting content into discussions, connecting users…

User experience

Access a bot

A bot is considered an Olvid “user” in its own right.

It is typically made available to other users via a Keycloak directory. Unlike the latter, a bot is a computer program running within your infrastructure, over which you have complete control.

A user can take advantage of the services provided by a bot by contacting it, just as he or she would with another user administered by the same directory. Once the bot has been added to the user’s contact list, he or she can “chat” with it directly, add it to a group discussion so that group members can take advantage of its services, introduce it to another contact, etc.

In all cases, the bot will have access to all messages sent in the discussion and will be able to insert messages.

Examples of services provided by a bot

A wide range of use cases are possible. Here are a few examples of bots we’ve tested internally or deployed with customers:

Information bots

The user receives messages from external sources directly in Olvid.

Visibrain integration

Visibrain is a social network monitoring tool, and the Olvid bot can receive alerts directly in Olvid. The communications team can then forward the most relevant alerts to the relevant users, directly in Olvid, via a simple message transfer.

Server room monitoring

The Olvid bot is connected to a temperature sensor and sends an alert message if the temperature rises. To make sure the recipient is aware of the situation, the bot can also send out a fake Olvid call to make the phone ring.

Crisis management

The Olvid bot is linked to a SOAR, and users are alerted in real time to critical security events. They can communicate directly with each other, in groups, to react to these messages and deal with incidents as quickly as possible. Discussion groups managed automatically by Keycloak ensure that the bot and the on-call team are always in the same group.

Mailing list

The bot is administered by a small group of administrators, who can send messages to the bot. The bot then broadcasts these messages to all Olvid users in the company with whom it is in contact. The company typically provides several bots of this type, on different themes. Users choose to invite the bots that are relevant to them, in order to “subscribe” to the mailing lists of their choice.

Chat bots

The user can interact with the bot via Olvid messages.

Private generative AI

The bot, hosted by the customer, is linked to an artificial intelligence model (such as Mistral) deployed on the same server. All users in contact with the bot can have a chat just as they would on the web with ChatGPT. The difference: all exchanges pass through Olvid, with a model deployed internally on the customer’s premises, so there’s no risk of data leaks.

Random animated gif

If one of the messages sent starts with !gif, the bot uses the rest of the message content to search an online gif platform (Giphy, Tenor, etc.) and posts the first gif found in the discussion.

Voice synthesis

The bot is connected to a loudspeaker and reads aloud all the messages it receives.

Business Bots

Users have easy access to business processes within the organization.

Help desk / Call center

A company’s customers are provided with a bot they can contact.This bot distributes Olvid license activation links to each new contact. All messages received by the bot are transferred to a ticketing system, where a help desk can answer customers’ questions. Customers can also place a call to the bot, which redirects the Olvid call to an available call center contact.

Connecting people

People outside an organization need to exchange sensitive information with experts inside the organization. These external people are registered in a web portal and can, via this portal, contact a bot which can “make the introductions” to put them in touch with the right expert for their needs. The exchange between the external person and the expert takes place entirely via Olvid, without the bot or web platform seeing the exchanges.

A law enforcement officer in the field takes a photo of a vehicle and sends it to the bot. The bot analyzes the image, extracts the license plate number and searches the central database. The search result is returned to the agent in a matter of seconds, directly into Olvid.

Position sharing tracking

Users in contact with the bot activate real-time position sharing. The bot receives updated positions from these users, and uploads this information to a platform that makes it easy to visualize the location of all users concerned on a map. The bot also keeps a history so that it can export the journeys of individual users (in GPX format, for example) for later analysis.

How it works

From a technical point of view, a bot is made up of two parts:

  • A daemon part that runs either in Docker or on a Linux server and exposes a gRPC API. This daemon implements a complete Olvid client (multi-profile in particular), based on the Linux version of Olvid.

  • A bot component, based on our Python library for exchanges with the daemon, which implements all the bot’s logic. This component can also run in Docker or on a Linux server (which can be the same server as the daemon). In this Python part, you can implement whatever you like to make a stand-alone bot or to interface it with your existing tools.