AI Prompt Node

Definitions

AI (Artificial Intelligence) is the simulation of human intelligence processes by machines, especially computer systems, to perform tasks such as learning, reasoning, problem-solving, and decision-making.

 

Performance impact

There are literally thousands of ways in which AI services can be added to business processes (digitised in Softools) to deliver steps changes in performance improvement.

 

6 typical example uses include:

- DATA QUALITY: Address issues of data completeness and quality in legacy systems / Excel

- DOC DATA EXTRACTION: Extract summary data from a linked / uploaded Document

- DATA KEY WORDS & CLUSTERING: Identify common labels for data records to accelerate search and retrieval

- COMPARATIVE ANALYSIS / MATCHING: Identify database records similar to the current record & situation

- DATA CREATION / IDEAS SUGGESTIONS: Generate solutions for common business issues (PromptAI/GenAI)

- SUMMARIES: Create summary reports for Executives based on multiple records

 

Getting going

All AI services are set up in the Workflow section:

 

As with other workflow items, there are 3 steps to follow:

- Trigger: what will initiate the workflow / AI service?

- Filter: what condition will be applied?

- Action: which AI service will be applied?

 

 

As more no-code AI Nodes are added to the Workflow options they will appear in the list of possible actions. OpenAI prompt is the most common use of AI at this time and is enormously flexible in terms of use cases.

 

 

Prompt engineering is the process of designing, optimizing, and refining inputs (prompts) given to AI models, such as language models, to achieve desired outputs or behaviour effectively and efficiently.

 

Prompt engineering is used across a wide range of applications to optimize how AI models generate outputs. Here are typical use cases:

 

1. Content Generation

· Writing articles, blog posts, or essays.

· Creating marketing copy or product descriptions.

· Generating creative content like poetry, stories, or scripts.

2. Text Summarization

· Summarizing long documents, articles, or reports.

· Extracting key points from meetings or research papers.

3. Data Analysis and Querying

· Generating insights from structured and unstructured data.

· Creating prompts to simplify complex data queries.

4. Language Translation

· Providing accurate and context-aware translations.

· Preserving tone, cultural nuances, and technical language.

5. Learning and Tutoring

· Designing educational content and personalized quizzes.

· Explaining complex topics in simpler terms.

6. Problem Solving and Ideation

· Brainstorming ideas for projects or businesses.

· Providing solutions to technical or strategic problems.

7. Sentiment Analysis

· Analysing and categorizing sentiment in text, such as customer feedback or social media posts.

8. Personalized Recommendations

· Generating tailored recommendations for books, movies, or travel plans.

· Creating prompts for e-commerce product suggestions.

9. Testing AI Models

· Evaluating and improving the model's behaviour with edge cases.

· Designing prompts to test specific functionalities or biases.

10. Legal and Compliance Support

· Drafting contracts, agreements, or legal clauses.

· Verifying compliance with specific regulations or policies.

11. Decision Support

· Generating pros and cons for decision-making.

· Analysing scenarios and providing risk assessments.

12. Customer Support

· Crafting templates for frequently asked questions.

· Escalating and solving complex customer issues via AI-powered responses.

 

Each use case requires carefully crafted prompts to maximize the relevance, accuracy, and efficiency of the AI model’s responses.

 

Creating your prompt

Once you have selected ‘Send Azure OpenAI Prompt’ you will need to fill in the configuration screen:

 

 

Node Type: This confirms that you are creating a node based on sending information to and receiving from Azure OpenAI service.

 

AI Service (Destination URL, API Version & API Key): Which AI service will you be using for this AI workflow. Your IT team will need to provide you with this information.

 

Max Tokens: Running AI services costs money based on the number of ‘Tokens’ consumed (based on the number of characters sent to and from the AI service). This limit will help to ensure your business does not spend more than intended on this service.

 

Temperature: The temperature setting determines how "creative" or "focused" the output should be, influencing whether the model generates diverse or more predictable outputs. Low (more predictable) 0.0-0.4, High (more balanced) 0.5-1.0

 

Response variable name: Set a variable name to store the prompt response choice message content. You will then refer to this variable in the next section.

 

Messages: This is where you define your prompt – the task you will be asking your AI service to perform.

In this example message the App Builder (User) is asking the AI service to calculate the financial cost saving based on data contained in another field. They have used the System and Assistant roles to give an example question and answer in order to improve the accuracy of response.

 

As you build up messages, use the 3 different roles to refine your prompt:

 

- System: instruct the purpose of the chat assistant e.g. ‘You are the data analyst in a financial services centre’

- User: example question e.g. ‘What are the top f trends in the storage of financial services data?’

- Assistant: best practice response to the example user question e.g. ‘The top 5 trends in the UK financial services market are: 1) ###, 2) ###,3) ###, 4) ###, 5 ###’

 

Notes: use this final section to capture your thinking. Sometimes the prompt can get very complicated and capturing your intent and logic will help ithers understand how this node works.

 

 

ACTION

Once the AI service has run, and your response have been received into the ‘Response Variable’, what do you want to do with it?

 

The most frequent use will be to use the ‘update field value’ option to enter the data into a field in your record.

 

____________________________________________________________________________________________________

CONTROL

 

Pre-deployment

- Before deploying your AI service on a production environment test it:

- Does it actually run / work?

- Does it give the quality of results you need?

- Does it hallucinate – make up values?

- Does it cost too much to run?

 

Post-deployment

Because AI services are changing so rapidly we recommend that you schedule a review of all AI services every quarter.

 

- New AI services are available that offer better results

- The cost of the service may have changed making some AI services too expensive or others now cheap enough to deliver a clear return on investment

 

______________________________________________________________________________________________________________

TIPS

Here are some general guidelines:

 

- Add multiple messages to the make the required response clearer

- Prompt Engineering is a combination of art and science - accept that every time you run the AI workflow it may create a different answer!

- Use trial and error to adapt your messages until you get consistent results in the format you need.

- Specify exactly what you want: e.g., add a User Message to tell the AI service to add the URL for any sources it provides, or ask it to provide the results in plain text or perhaps in Japanese.

- Be explicit with the expected response e.g. System - "Only return the percentage value, without % symbol"

- Users can access the full json response, also useful in SendWebhook node e.g. jsonvalue([Meta.ResponseContent], 'choices[0].message.content', 'text')