https://play.orkes.io/workflowDef/mental_health_emergency_response
This project aims to create a comprehensive workflow that addresses mental health emergencies by combining speech-to-text technology with emotional analysis and first responder alerts.
Key Objectives:
1.Speech-to-Text Implementation:
Set up the openAI Speech-to-Text API and test it with sample audio files. Create a task in Orkes that sends audio data to the openAI API and captures the transcribed text.
2.Emotional Analysis:
Use Hugging Face’s NLP models to detect emotions from the transcribed text or mock this process with predefined responses. Implement a task to analyze the emotional content and provide stress management responses based on detected emotions.
3.Responder Connection:
Set up a mock API (e.g., Twilio) to alert first responders about the user's emotional state and location, ensuring timely assistance.
4.Workflow Design & Testing:
Design the entire workflow in Orkes, linking all tasks in a logical sequence. Conduct thorough testing of each individual task and then run end-to-end tests to validate the workflow's effectiveness.