Add Low Token/Quota Notifications for AI Scribe to Prevent Mid-Consultation Interruptions

[dpnkr1]

AI Scribe in CARE HMIS operates on a token-based quota system. While administrators can view the available tokens from the Quota Details page, doctors are often unaware when their token allocation is nearing exhaustion.

During field usage, it has been observed that doctors may dictate lengthy clinical notes, only for the AI Scribe quota to run out in the middle of a consultation. This interruption is frustrating and disrupts the documentation workflow.

Problem Statement

1. Doctors do not receive any indication that their token quota is running low.
2. Token exhaustion during an active consultation interrupts the AI Scribe workflow.
3. Administrators may not notice low token balances until users start experiencing failures.
4. Unexpected interruptions lead to poor user experience and frustration.

Proposed Solution

Introduce proactive low-token alerts:

1. Generate warnings when the remaining token balance falls below configurable thresholds (e.g., 20%, 10%, or 5%).
2. Notify facility administrators via in-app notifications when the facility token pool is running low.
3. Optionally, notify users when their individual token/quota allocation is nearing exhaustion.
4. Display a warning before starting a transcription if the remaining quota is insufficient for a typical consultation.

Expected Outcome

1. Prevent AI Scribe interruptions during consultations.
2. Allow administrators to replenish quotas before they are exhausted.
3. Improve doctor satisfaction and trust in the AI Scribe feature.
4. Ensure smoother and uninterrupted clinical documentation workflows.

Additional Context

Although token availability can already be viewed on the Quota Details page for admin, users typically become aware of low quotas only after the AI Scribe stops functioning. Proactive notifications would help prevent disruptions and improve the overall user experience.

🚨 DO NOT EDIT BELOW THIS LINE 🚨

Instructions for Requesting Assignment:

To request assignment, please clearly outline your solution and timeline by commenting on the issue using the format below:

Describe your solution clearly:
Provide a detailed explanation of your proposed solution, including your approach, key implementation steps, and relevant examples or references. Mention any dependencies, assumptions, or risks you foresee that might affect your timeline or implementation.

Expected Timeline:

• End date: [Expected submission date of a completed Pull Request]

Additional Context:
Include any other relevant context, links, screenshots, or resources that support your proposed solution.

🚨 Your assignment may be unassigned if there is no activity or progress within the stated timeline unless communicated clearly and agreed upon.