Text Prompt Flow Endpoint Consumption

Farah Diana Masri 40 Reputation points
2024-12-30T00:58:17.9366667+00:00

Hi,

I want to ask questions about text prompt flow endpoint consumption. I have created a text prompt flow following this documentation. Then, I deploy this prompt flow so I can consume the endpoint to my own application. However, I just realized that my own application does not have the conversation history. I already tested the endpoint in Azure AI Foundry for the conversation history and it works but in my own application it does not work. This is my code for back-end. I have tried to declare the conversation history alone in my code, but it seems like, it follows the input in the text prompt flow I created before. Kindly request help on how to enable the conversation history in this code or do I need to modify the front-end code also?
Thank you so much for helping.


import axios from 'axios';

const apiKey = process.env.AZURE_KEY;
const endpoint = process.env.AZURE_ENDPOINT;

export const askQuestion = async (question) => {
    const data = {
        chat_history: [], // Initially empty, can be populated as needed
        chat_input: question
    };

    try {
        const response = await axios.post(endpoint, data, {
            headers: {
                'Content-Type': 'application/json',
                'Authorization': `Bearer ${apiKey}`,
                'azureml-model-deployment': 'bmi-ai-bishub-project-gpt4o-1'
            }
        });

        if (response.data) {
            // // Extract and return only the raw output
            return response.data.chat_output || JSON.stringify(response.data, null, 2);
        } else {
            throw new Error('No response from the server');
        }
    } catch (error) {
        throw new Error(error.response ? error.response.data : error.message);
    }
};

Azure AI services
Azure AI services
A group of Azure services, SDKs, and APIs designed to make apps more intelligent, engaging, and discoverable.
3,063 questions
{count} votes

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.