The AI Assistant is an AI-powered tool within the Canvas of the Digibee Integration Platform. As part of our AI Pair Programmer approach, it helps accelerate time-consuming tasks, such as:
Designing integration logic
Generating JOLT specifications
Getting general guidance and support
With the AI Assistant, you can complete these tasks without leaving the environment where you're building your integrations.
How to use the AI Assistant
Access the AI Assistant by clicking the AI icon in the left menu. Once opened, you will see a list of options tailored to your goal:
In any of these options, you can give feedback by liking or disliking the answer, helping us improve our AI models.
Scaffold a pipeline
Provide a prompt describing the logic of your integration, and receive a structured pipeline in return. This includes:
A flow tree
The integration steps
A summary of the structure
Once you receive the pipeline structure, you can:
Adjust the flow by asking the AI for changes until you achieve the desired result.
Copy the flow to use elsewhere.
Insert the flow directly into the Canvas of your pipeline.
Create JOLT spec from samples
Send the JSON input to the AI. If needed, you can make changes to the input. Once it's processed, provide the expected output. The AI will then return the corresponding JOLT specification.
After receiving the JOLT specification, you can:
If the answer isn’t what you expected, you can give the AI more context to refine the output until it meets your needs.
Get general help
Ask anything about Digibee and get an answer based on our Documentation Portal. This gives you fast access to the information you need to create, configure, and troubleshoot your integrations.
Writing effective prompts for AI-generated pipelines
When using AI to generate a pipeline, the way you describe your request matters. Providing clear details helps the AI understand your needs and generate an accurate pipeline structure. Below are examples of good and bad prompts to help you get better results.
Prompts with specific connectors
If you already know which connectors your integration needs, mention them in your prompt. This helps the AI generate a more accurate pipeline.
Do: Provide clear, step-by-step details.
✅ I want to create a pipeline that starts by using the REST V2 connector to make a request. After that, it will use a Choice connector to evaluate the API response, followed by a Log connector to check the result. If the request is successful, the pipeline should use the JSON Generator to create the output message. If there’s an error, the Email V2 connector will be used to send a notification, ending the flow.
Don’t: Be vague or general. Without connector names, the AI might not generate the correct flow.
❌ I want to create a pipeline that makes an API request, checks the response, and then sends an email if something goes wrong.
Why is this unclear? This doesn’t specify which connectors to use, what kind of API request, or what “something going wrong” means.
Prompts with endpoint types
If you don’t know the exact connectors, but know the types of endpoints involved, you can describe the integration at a high level and let the AI select the appropriate connectors.
Do: Mention the systems or endpoints types clearly.
✅ Connect my SAP ERP system to a database and a REST API.
Don’t: Be too generic. If you only say “ERP system”, the AI won’t know which kind you’re referring to.
❌ Connect my ERP system to a database and a REST API.
Why is this unclear? Different ERP systems might require different connectors. Providing the system name helps the AI choose the best options.
Prompts without mentioning connectors
If you are unsure which connectors to use, you can still write a detailed prompt that describes your goals. The AI will then determine the best approach.
Do: Clearly describe the steps of the process.
✅ I need to create a pipeline integration that reads data from a stream database, publishes an event, and writes the data to another database. It starts by connecting to the stream database to read the necessary data. The data is then published as an event to notify other systems or services. Finally, the data is written to another database for storage or further processing.
Don’t: Leave out key details. A vague request might result in an incomplete or incorrect pipeline.
❌ I need a pipeline that reads data, publishes it, and writes it somewhere else.
Why is it unclear? It doesn’t specify what kind of data, where it’s coming from, what kind of event needs to be published, or what database is involved.
Additional tips
Keep these in mind when formulating your prompt:
Be as specific as possible about the structure of the flow you want to generate.
Write clear, direct instructions describing the part or the entire pipeline you want the AI to generate.
Mention specific connectors if you already know which ones to use.
If your prompt is complex, break it into parts and ask the AI to make adjustments if needed.
FAQs
Can I create large pipelines using AI?
Yes, but the AI may generate hallucinations when handling excessively large pipelines.
Does the AI understand the existing pipeline on the Canvas?
No, the AI doesn’t recognize the context of an existing pipeline on the Canvas. It generates new content without associating it with the current pipeline.
Does the AI configure the connector’s parameters?
No, the AI doesn’t set connector parameters. Its main focus is on defining the logical structure and connections between connectors.
Can I use the AI Assistant for both pipelines and capsules?
Is the chat history saved in the AI Assistant?
Yes. Chat history is saved per pipeline and per user. Every time you access a pipeline where you previously interacted with the AI Assistant, you will be able to retrieve the chat history for that pipeline.
See below for a .
Copy it and paste it into the connector.
OpenAI’s documentation provides a comprehensive guide on creating more precise prompts through prompt engineering. .
No, for now, the AI Assistant is only available for pipelines. However, some AI-powered features are available for capsules. Learn more about the .
