Quantum Minds Flow Operators

Introduction

Flow operators in Quantum Minds are essential for creating dynamic, adaptive minds that can make decisions and control execution paths based on data, conditions, and user inputs. These operators enable you to build sophisticated workflows that respond intelligently to different scenarios, making your minds more powerful and flexible.

Available Flow Operators

Flow.Condition (IfElseThen)

The Flow.Condition operator (formerly IfElseThen) evaluates conditions and directs the flow of execution to different paths based on the result, enabling decision-making capabilities within your minds.

Inputs

Outputs

Example Usage

Prompt: "Check if the quarterly sales figures exceed the target of $500,000"

Output:
- type: "markdown"
- then_: "Path to follow if sales > $500,000"
- else_: "Path to follow if sales <= $500,000"

Condition Evaluation Methods

The Flow.Condition operator supports multiple approaches to condition evaluation:

1. Direct Condition Checking

Using simple comparisons and logic:

Prompt: "Is the average customer satisfaction score from $CustomerFeedback_001.output.content greater than 4.2?"

2. Data-Based Decisions

Evaluating based on data analysis:

Prompt: "Based on the sales trend data in $SQLExecution_001.output.content, determine if there is a statistically significant upward trend (p-value < 0.05)"

3. Complex Logic with Multiple Factors

Combining multiple conditions:

Prompt: "Evaluate if ALL of the following conditions are met:
1. Monthly revenue exceeds $100,000
2. Customer acquisition cost is below $50
3. Churn rate is less than 5%
Based on data in $Analytics_001.output.content"

4. Natural Language Understanding

Processing conditions expressed in natural language:

Prompt: "Is the customer message in $CustomerMessage_001.output.content expressing dissatisfaction or requesting a refund?"

Creating Flow Control Structures

Basic Branching

The simplest use of Flow.Condition creates a two-way branch:

{
  "operator": "PandasAi",
  "input": {
    "prompt": "Analyze customer churn data and calculate the overall churn rate",
    "dataframe": "$SQLExecution_001.output.content"
  }
}

↓

{
  "operator": "Flow.Condition",
  "input": {
    "prompt": "Is the churn rate calculated in $PandasAi_001.output.content greater than 5%?"
  }
}

then_ →

{
  "operator": "TextSummarize",
  "input": {
    "prompt": "Generate an urgent action plan to address the high churn rate of over 5%"
  }
}

else_ →

{
  "operator": "TextSummarize",
  "input": {
    "prompt": "Generate a standard churn monitoring report for the acceptable churn rate"
  }
}

Multi-Level Conditionals

Flow conditions can be nested to create more complex decision trees:

{
  "operator": "Flow.Condition",
  "input": {
    "prompt": "Is the sentiment in $CustomerFeedback_001.output.content positive?"
  }
}

then_ →

{
  "operator": "Flow.Condition",
  "input": {
    "prompt": "Does the feedback in $CustomerFeedback_001.output.content mention our new product features?"
  }
}

then_ →

{
  "operator": "TextSummarize",
  "input": {
    "prompt": "Generate a response thanking the customer for positive feedback on our new features"
  }
}

else_ →

{
  "operator": "TextSummarize",
  "input": {
    "prompt": "Generate a general thank you response for positive feedback"
  }
}

else_ →

{
  "operator": "Flow.Condition",
  "input": {
    "prompt": "Is the sentiment strongly negative (rating 1-2 out of 5)?"
  }
}

// Additional branches...

Implementing Loops with Recursion

While there's no explicit loop operator, you can implement iterative processing through recursion and conditions:

{
  "operator": "TextToSQL",
  "input": {
    "prompt": "Get the next batch of 100 unprocessed orders",
    "dataset": "order_management"
  }
}

↓

{
  "operator": "SQLExecution",
  "input": {
    "sql": "$TextToSQL_001.output.content",
    "dataset": "order_management"
  }
}

↓

{
  "operator": "Flow.Condition",
  "input": {
    "prompt": "Does the result in $SQLExecution_001.output.content contain any orders (rows > 0)?"
  }
}

then_ →

{
  "operator": "PandasAiDf",
  "input": {
    "prompt": "Process each order...",
    "dataframe": "$SQLExecution_001.output.content"
  }
}

// Continue with saving results and then loop back to the beginning

else_ →

{
  "operator": "TextSummarize",
  "input": {
    "prompt": "Generate processing complete summary"
  }
}

Advanced Flow Control Patterns

Error Handling

Use Flow.Condition to implement error detection and recovery:

{
  "operator": "CustomPython",
  "input": {
    "code": "try:\n    # Complex operation\n    result = process_data()\n    error = None\nexcept Exception as e:\n    result = None\n    error = str(e)"
  }
}

↓

{
  "operator": "Flow.Condition",
  "input": {
    "prompt": "Check if error is None in $CustomPython_001.output.result"
  }
}

then_ →

{
  "operator": "TextSummarize",
  "input": {
    "prompt": "Continue with normal processing..."
  }
}

else_ →

{
  "operator": "TextSummarize",
  "input": {
    "prompt": "Generate error report and recovery plan based on: $CustomPython_001