🏥 Fluxo - AI Healthcare Voice Assistant

ASEAN AI Code Fest 2024 - Supercharged Clinician Track

🎯 Project Overview

Fluxo is a revolutionary AI-powered voice recording and medical documentation system designed to address clinician burnout in Malaysia’s healthcare system. This application transforms how healthcare professionals capture, transcribe, and analyze patient consultations through intelligent automation.

🚀 Problem Statement

Malaysia’s post-COVID healthcare system faces a perfect storm:

• 74% of government healthcare workers burned out and 73% considering quitting
• Nearly half of a doctor’s workday spent on paperwork instead of patients
• Duplicate data entry across siloed systems, increasing error risk
• High patient load: up to 25–30 patients per doctor per session, with 3-minute consultations
• This leads to rushed care, missed critical details, and clinician attrition. The current “business as usual” approach is unsustainable.

💡 Proposed Solution

Fluxo is an AI-powered medical scribe designed for Malaysian clinicians, especially in public hospitals. It:

• Listens to doctor-patient consultations in real time
• Generates MOH-compliant SOAP notes instantly in English & Bahasa Malaysia
• Flags red-flag symptoms and missing details
• Integrates with existing hospital EHR to eliminate duplicate entry
• Summarizes patient history from disparate data sources into one view

📊 Measurable Impact

• 35% - 50% reduction in documentation time + reduce after office work (catchup)
• Improved clinical accuracy through AI validation
• Enhanced patient care via automated insights
• Scalable solution for Malaysia’s healthcare system

🎨 User Experience

Desktop Medical Console

• 3-Panel Layout: Patient list, recording interface, medical notes
• Professional Design: Clinical-grade UI with medical iconography
• Real-time Features: Live waveform visualization and transcription
• Medical Documentation: Structured SOAP notes with audio transcribtion and medical notes

Mobile Healthcare Experience

• Touch-optimized Interface: Large buttons and gesture-friendly controls
• Adaptive Navigation: Single-panel view with bottom tab navigation
• Seamless Workflow: Smooth transitions between recording, notes, and patient list

🛠️ Technology Stack

Core Technologies

• Frontend: Next.js 15.4.6, React 19.1.1, TypeScript 5.9.2
• Styling: Tailwind CSS 3.4.17, shadcn/ui components, Framer Motion 12.23.12
• Audio: Soniox Speech-to-Text Web SDK, Web Audio API, MediaRecorder API

Development Libraries

• UI Components: Radix UI primitives (@radix-ui/react-avatar, @radix-ui/react-dialog)
• State: React hooks, Zustand 5.0.7, TanStack Query 5.84.2
• Validation: Zod 4.0.16, React Hook Form 7.62.0
• Animation: Framer Motion 12.23.12, Tailwind CSS Animate
• Build Tools: ESLint, TypeScript compiler, PostCSS, Autoprefixer

APIs & Services

• Speech Recognition: Soniox Speech-to-Text API with speaker diarization
• Audio Processing: Web Audio API for real-time capture and analysis
• AI Integration: Ready for OpenAI GPT integration for medical analysis

🏥 Healthcare-Specific Features

Medical Documentation (SOAP Format)

• Subjective: Chief complaint and patient history
• Objective: Vital signs and examination findings
• Assessment: Clinical diagnosis and evaluation
• Plan: Treatment, medications, and follow-up instructions

Clinical Workflow Integration

• Patient Information Management: Avatar, visit types, consultation history
• Session Notes: Real-time note-taking during consultations
• Medical Entity Extraction: Symptoms, medications, conditions
• Timestamp Tracking: Precise recording and documentation times

Datasets:

MOH public guidelines, anonymized SOAP notes, WHO medical terminology

Prompt Engineering:

Structured SOAP generation, red-flag symptom highlighting, bilingual translation

Scalability

Short-term: Deploy Fluxo in MOH hospitals for high-burnout specialties such as internal medicine, emergency care, and oncology.

Mid-term: Leverage our AI agent architecture to rapidly adapt Fluxo for additional medical specialties (e.g., pediatrics, obstetrics, psychiatry) by fine-tuning specialty-specific prompts, terminology sets, and clinical workflows — without major code changes.

Long-term: Expand nationwide and into ASEAN markets, with multilingual capabilities (e.g., Thai, Vietnamese, Tagalog) and local regulatory compliance. Fluxo’s modular AI agent design ensures specialty modules can be added in weeks, not months, enabling fast scaling to diverse healthcare environments.