Service model and fees

Subscription model: AI Agent as a Service

You pay for a monthly subscription and we handle everything:
The creation of the agent, its integration and maintenance under a service-delivery agreement. If you're not satisfied, you cancel at the end of any month

Ownership and control:

Initial consultation:

We seek to understand the client's problems, objectives and infrastructure

Assessment:

We analyse the clients needs, research options and evaluate solutions

Proposal:

We explain our solution, its value proposition, costs and timeline

Implementation:

Development, deployment and quality assurance

Roll-out:

Ensuring easy adoption with minimal disruption to existing workflows

Follow up and support:

We accompany the client and make sure everything works  

Ownership model: Build & release

Once the evaluation and testing phases are successfully completed, you pay a one-off invoice reflecting what was agreed in the software-development contract

Ownership, control and maintenance:
  • The agent and its integrations belong to you and are hosted in the cloud in accounts under your name and under your control
  • The data is yours and you manage it
  • iAutomator retains no ongoing access unless you hire us for maintenance work, granting us access under the principle of least privilege
Automator remains open to negotiating hybrid ownership and billing models between those outlined above

How much does an iAutomator AI agent cost?

Each agent is unique and designed for each company; your agent is built around your specific needs and the cost depends on several factors:
Functionality: The number and complexity of features is critical

Documentation:
If the information is already organised and in formats easily processed by AI, the cost is much lower than when it is disorganised, requires preprocessing or recognition steps

Internal integrations:
If your systems are modern they will be easier to connect than legacy systems, which require more development and maintenance. In the case of very old systems, integration may not be possible

External integrations:
If there is a native connection, integration is simple; if not, it depends on well-documented MCP servers or APIs

Platforms, tools and models:
Prices vary according to the model, the platform and the vector databases used