A full set of features for collaborative IoT

By using fast and precise natural language processing (NLP) and enhanced by decision-making skills, connected things can integrate workflows that combines other things, services and people on a any-to-any exchange platform.
Natural language at the speed of machine language

Machine-to-machine communication requires high precision (no action should be taken wrongly), well above the usual performance of human-to-machine services.

Without human in the loop, there is no trivial possibility to correct erroneous interpretations, ask for confirmation or enter a repair phrase ("can you confirm?”, “Which color would you prefer?”). Therefore, most dialogues must be concluded in one round only.

Natural language interactions with machines appear as quickly as with machine language, that is, only a few milliseconds.

CPU workload and memory requirements are kept very low to maximize scalability and minimize operation costs.

High precision and concision

Ultra fast and lightweight

The three steps toward IoT collaboration

The first step is to provide connected devices with the ability to communicate in natural language.

Then, augmented capabilities associated to each thing (rich events handling, workflow orchestration, etc.) provide the meaningful content to be exchanged.

Finally, these natural language messages are shared by the different members of the workgroup (devices, applications, people).

Augmented capabilities

Learning to speak is not enough: connected things must have something interesting to say. Their common (natural) language must be used to communicate significant information (commands, status, values) that support their mission: a "soul" must be breathed into these speaking devices.


Rich events handling

Connected devices are capable of generating events (for example "temperature above 22 °C"). Agora's vThings can also generate their own events to enrich real devices' capabilities.


Workflows orchestration

Talking devices are complemented by distributed orchestration capabilities that implement event-driven workflows involving heterogeneous things, applications and people.

Any-to-any communication

Participants are brought together in a secure environment (workspace) which offers confidentiality from the outside and any-to-any communication inside.

Rich events handling to create complex states and useful trigger conditions

Embedded orchestration to support business oriented workflows

Any-to-any communication inside a secure workspace organization

Main features

Ultra fast and precise bidirectional traduction between natural language and machine language used by each connected devices and applications

Natural language/machine language adaptation

Workspaces can be connected to the most popular social networks so users can interact with devices and services from their usual tools

Connection to public
social networks

Workspaces can be connected to the most popular enterprise collaboration platforms so that your teams can collaborate with devices and services in natural language

Connection to enterprise collaboration platforms

Workspaces can be connected to enterprise productivity tools like calendars, sheets and data analytics for a smooth integration

Connection to enterprise productivity tools

The complete Verbatim provides natural language tracability of all communications within workspaces (users, devices, applications)

Natural language Verbatim for direct traceability

Business analysts can define event-driven workflows combining connected devices and applications without a line of code

No-code event-driven workflow controller

Bundling of devices and services facilitates communications and workflows for large-scale deployments

Devices and service grouping

Users can communicate in their preferred natural language with their connected devices and services

Multi-language user interactions

Workspace administrators can simply select pre-existing IoT and application adapters (vThing) from a regularly populated store

Access to a large store of service interfaces (vThings)