FocusFlow is a bio-adaptive software ecosystem designed to eliminate cognitive exclusion by dynamically remapping the digital world to fit the human brain. The platform serves as an “external frontal lobe,” utilizing Azure AI and multimodal processing to transform overwhelming data into structured, high-signal intelligence in real time.
Modern digital environments are built on cognitive uniformity, assuming every user possesses identical executive function capabilities.
For individuals with ADHD or dyslexia, the digital world becomes a source of:
- “Wall of Text” anxiety
- Auditory processing delays
- Task paralysis
Standard learning tools fail to address the high friction caused by constant switching between auditory, visual, and textual contexts—leading to rapid cognitive fatigue and disengagement.
FocusFlow bridges the executive function gap through a dual-interface system:
A centralized hub for deep processing of static assets:
- PDFs
- Long-form text
- Links
- Downloaded videos
A real-time intervention layer for:
- Live meetings (Google Meet / Microsoft Teams)
- Web navigation
Provides active recall assistance and visual logic mapping on the fly.
- Quick content summarization
- PDFs, long texts, links, and videos transformed into:
- Bionic text
- Visual diagrams
- Flowcharts
- Flashcards
- Optimized for cognitive understanding and long-term retention
-
Live Bionic Transcription
Real-time audio-to-context conversion using Azure Speech Services with Bionic anchors to prevent tracking loss. -
Smart Nuggets
AI-driven summarization prioritizing:- Definitions
- Actionable tasks
Using color-coded semantic tags.
-
Topic Segmentation
Azure Video Indexer–powered color-coded navigation across meeting or video timelines. -
The Shredder
Converts dense transcripts and slides into structured flowcharts to visualize logic paths. -
Visual Analysis (OCR Snipper + Chat)
Azure AI Vision-powered tool that:- Captures slide or video regions
- Extracts handwriting or text
- Passes it to a specialized visual chatbot
-
Output Diversity Export content as:
- Bionic text
- Flowcharts
- Visual diagrams
- Flashcards
-
Bionic Web Engine
Instantly converts any website (e.g., Wikipedia) into Bionic format. -
Contextual Chat
ADHD-friendly chatbot that:- Explains selected text
- Supports follow-up questions
- Adapts explanations visually and structurally
-
Non-Invasive Focus Tracking
Computes a real-time Focus Score (0–100) using:- Tab focus
- Idle thresholds
- Scroll behavior
(No webcams, no keystrokes)
-
AI Companion
- Monitors focus state
- Politely alerts when attention drops
- Triggers micro-quizzes based on current content to reinforce retention
FocusFlow is built on a high-concurrency monorepo architecture, ensuring sub-second latency for real-time AI interventions.
-
Core Logic
Node.js + TypeScript environment handling complex data remapping. -
AI Processors (
/processors)
Dedicated workers for:- Flashcard generation
- Flowchart construction
- Bionic summarization
-
Identity & Progress (
/controllers)
Manages:- User bio-preferences
- Study streaks
- Focus session analytics
-
Dashboard (
/pages/Dashboard.tsx)
High-performance React interface using progressive disclosure to reduce cognitive overload. -
Bionic Engine (
/components/dashboard/summaryrender.tsx)
Custom Markdown parser converting raw content into visual Bionic anchors. -
Extension Logic (
/content/vision.js)
Handles:- Canvas-level OCR capture
- Real-time DOM manipulation for Bionic remapping
| Azure Service | Role |
|---|---|
| Azure Cosmos DB | Stores user preferences, bio-adaptive settings, and session metadata |
| Azure Blob Storage | Durable storage for processed JSON assets and visual exports |
| Azure AI Foundry | Custom-tuned ADHD-optimized summarization & recall models |
| Azure AI Vision | Real-time OCR and image analysis |
| Azure Video Indexer | Multimodal topic segmentation & scene detection |
| Azure Speech Service | Low-latency diarization and transcription |
FocusFlow integrates seamlessly into existing educational and corporate workflows.
The browser extension enables zero platform modification, working instantly with:
- Google Meet
- YouTube
- Internal LMS systems
The MVP delivers immediate value through:
- Live audio-to-text transcription with ADHD-friendly formatting
- Bionic summarization for text, links, and PDFs
- Rule-based Focus State Machine:
FOCUSEDAT_RISKLOST
Supportive nudges replace disruptive alerts.
-
Azure Immersive Reader Integration
Enhanced accessibility and voice-assisted reading. -
Bio-Feedback Focus Loop
Optional gaze tracking (Azure Vision) to pause sessions during focus gaps. -
Study Companion Agent
- Proactively suggests breaks
- Detects erratic mouse movement or digital “fidgeting” patterns
-
Educational
Reduces Time-to-Understanding for neurodivergent learners. -
Social
Promotes cognitive equity in remote education and workspaces. -
Privacy
Privacy-first design with local bio-data processing for high-trust adoption.
- Node.js v18+
- Azure Subscription (Cosmos DB, Blob Storage, AI Services)
- Gemini 1.5 Pro / Flash API Key (via Azure AI Foundry)
- Clone the repository
- Configure
.envfiles in:apps/web/serverapps/extensions/backend
- Run
npm installat the monorepo root - Launch locally using:
docker-compose up