Implement Conditional Logic in AI Flows
Implement conditional logic in your AI workflows with this prompt flow example for if-else scenarios.
Why it matters
Orchestrate AI processes with dynamic conditional branching. This asset enables your AI to make decisions and execute different paths based on input criteria, mimicking if-else logic.
Outcomes
What it gets done
Define conditions for AI process branching
Execute specific AI tasks based on evaluated conditions
Create flexible and responsive AI workflows
Integrate decision-making into automated processes
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/pf-standard-conditional-flow-for-if-else | bash Capabilities
What this chain does
Labels or categorizes text, files, or data points.
Pulls structured data fields from unstructured text.
Condenses long documents or threads into key takeaways.
Overview
Conditional Flow For If Else
What it does
This prompt flow example demonstrates how to implement conditional logic within a multi-step AI workflow. It is designed for scenarios requiring an if-else structure, allowing the flow to take different execution paths based on specified conditions.
How it connects
Use this prompt flow when you need to create AI applications that can dynamically alter their behavior or output based on whether a certain condition is met. It's ideal for scenarios requiring branching logic in your prompt engineering efforts.
Source README
Conditional flow for if-else scenario
This example is a conditional flow for if-else scenario.
By following this example, you will learn how to create a conditional flow using the activate config.
Flow description
In this flow, it checks if an input query passes content safety check. If it's denied, we'll return a default response; otherwise, we'll call LLM to get a response and then summarize the final results.
The following are two execution situations of this flow:
if input query passes content safety check:
else:
Notice: The content_safety_check and llm_result node in this flow are dummy nodes that do not actually use the conten safety tool and LLM tool. You can replace them with the real ones. Learn more: LLM Tool
Prerequisites
Install promptflow sdk and other dependencies:
pip install -r requirements.txt
Run flow
- Test flow
# test with default input value in flow.dag.yaml
pf flow test --flow .
# test with flow inputs
pf flow test --flow . --inputs question="What is Prompt flow?"
- Create run with multiple lines of data
# create a random run name
run_name="conditional_flow_for_if_else_"$(openssl rand -hex 12)
# create run
pf run create --flow . --data ./data.jsonl --column-mapping question='${data.question}' --stream --name $run_name
- List and show run metadata
# list created run
pf run list
# show specific run detail
pf run show --name $run_name
# show output
pf run show-details --name $run_name
# visualize run in browser
pf run visualize --name $run_name
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.