FitCare Wellness - Dietician Virtual Assistant (Agentra)

Executive Summary

Problem Context

FitCare Wellness offered nutrition plans but faced a bottleneck — dieticians were spending hours answering repetitive dietary queries, delaying new patient onboarding and reducing consultation slots.

• Manual consultation booking & follow-ups slowed patient onboarding.
• Dieticians spent 40% of their time on repetitive Q&A instead of high-value consultations.
• Patients had to wait 2–3 days for diet plan adjustments.

Solution: Agentra AI Agent Deployed

Agentra Talk + Schedule Agent

Automated nutrition consultation, follow-ups, and appointment scheduling.

Standalone, Web Widget, Voice, Avatar on Healthcare Website

1. Collects patient dietary goals, allergies, and preferences via chat/voice.
2. Uses LLM to interpret health requirements and match with nutrition guidelines.
3. Pulls recipes and meal plans from the clinic's internal database via RAG.
4. Books follow-up consultations directly in the clinic calendar.
5. Escalates to a live dietician only for complex cases.

Agent Architecture View (Agentic Workflow)

• LLM: GPT-4 via LangChain for reasoning & conversation
• Memory: Remembers patient history & preferences
• RAG: Fetches meal plans from internal nutrition database
• Tools: Google Calendar, EMR Integration
• Planner: Multi-step decision orchestration (gather info → suggest plan → schedule → follow-up)

Business Impact & Outcomes

• Wait time for dietary adjustments: 2–3 days → under 15 minutes
• Reduced manual Q&A load on dieticians by 70%
• Patient onboarding capacity increased by 35%

Key Features Showcased

• Multi-modal AI: Patients could talk to the AI via text, voice, or see an avatar.
• Agentic Tool Use: Automated bookings in Google Calendar & EMR updates.
• LLM Reasoning: Provided contextual diet recommendations based on history.
• Autonomy: End-to-end handling from intake to follow-up with minimal handover.
• Memory & Personalization: Suggested meal plans based on past interactions.

Agent in Action (Workflow)

1. Patient starts a voice/text request
2. Agent clarifies hypertension symptoms
3. Agent retrieves & explains a suitable meal plan
4. Books a follow-up in the calendar
5. Sends daily reminders for adherence

Why Agentra?

• Pre-built healthcare-ready templates for quick deployment.
• Multi-channel engagement without separate integrations.
• Strong API & EMR connectivity for healthcare compliance.
• Agentic workflows allowing autonomous, multi-step patient handling.