Skip to main content

Scaling and Customizing

You can further scale the process by adding more steps and using different models. For instance:

  • Use GPT-4o mini for simpler tasks to reduce costs.
  • Integrate retrieval-augmented generation for more accurate responses.
  • Incorporate vector databases or custom functions for enhanced orchestration.

The flexibility of entAIngine allows you to mix and match models and workflows to create the most efficient process for your specific needs.

Finalizing and Automating the Workflow​

To fully automate the workflow, consider adding further steps beyond just drafting an email. For example:

  • Approval Step: Once the email is sent, the responsible person can approve or decline the payment. Set up another task to automatically update the CFO on the response.
  • Data Logging: Store all processed invoices and responses in a central database for tracking and auditing purposes.
  • Notifications: Implement automatic notifications to alert relevant stakeholders about the status of each invoice. For example, if a response is delayed, a follow-up reminder could be triggered.

These steps help streamline the process from start to finish, ensuring that the entire invoice approval cycle is handled with minimal manual intervention.

Example Application Types in entAIngine​

entAIngine offers various application types that can be used for different purposes:

  • AI Process Application: A core application used to automate workflows by orchestrating multiple AI services.
  • Company Guru Application: Works like a company-specific knowledge repository, allowing users to retrieve information quickly without needing a traditional chat interface.
  • Chat Application: An interactive interface ideal for tasks that require real-time engagement.
  • Editor Application: Creates documents with embedded AI snippets, perfect for iterative tasks that involve a human-in-the-loop approach.

For our example, we used an AI Process Application. Depending on the complexity of your business needs, you might explore other application types offered by entAIngine.

Summary​

By following these steps, you have built your first multi-step process in entAIngine. This guide demonstrated how to automate a finance workflow by breaking it down into smaller tasks and using appropriate models to complete each step. Additionally, we explored ways to scale, customize, and extend the process to achieve a fully automated and efficient workflow.

By adopting this divide-and-conquer approach, you can create reusable, efficient processes that save time, ensure consistent quality, and provide real value across your organization.

Next Steps​

Now that you have successfully created your first process, consider exploring more complex use cases. Use entAIngine's extensive library of AI models and connectors to build applications tailored to your organization's specific needs. Check out our detailed guides on advanced features like retrieval-augmented generation, integrating with external APIs, and more to take your automation capabilities to the next level.