Daily Knowledge Tracker
Development Status & Roadmap
Project Overview
"Empowering parents to take control of their child's academic journey through AI-driven personalization."
The Motive: Daily Knowledge Tracker (DKT) was built to bridge the gap between classroom learning and home practice. Recognizing that every student learns at a different pace, DKT leverages advanced AI to generate infinite, curriculum-aligned practice papers tailored specifically to a child's weak areas.
The platform empowers parentsโwhether technically savvy or notโto act as effective tutors. By providing instant marking, detailed explanations, and visual progress tracking, we ensure consistency in learning without the massive overhead of manual paper creation or expensive tuition.
๐Development Journey
Phase 1: Foundation
COMPLETED- โ Authentication (OAuth)
- โ Database Schema Design
- โ Basic LLM Integration
- โ Admin Panel Core
Phase 2: Core Features
COMPLETED- โ PDF Generation Engine
- โ Online Marking System
- โ Analytics & Charts
- โ Multi-Student Support
Phase 3: Polish
COMPLETED- โ UI/UX Overhaul
- โ Global Font System
- โ Build Optimization
- โ Performance Tuning
Phase 4.1: Access Controls
COMPLETED- โ Max 5 Students Limit
- โ 14-Day Trial Logic
- โ Read-Only Mode
- โ Admin Overrides
Phase 4.2: Quality & Discipline
IN PROGRESS- โ Pedagogy (Year Group Rules)
- โ Deduplication (Anti-Spam)
- โ Dynamic Duration
- โ Mark-Before-Generate (Pending)
Phase 4.3: Analytics
PLANNED- โ Consistency Benchmarks
- โ Velocity Charts
- โ Target: 20 Papers/Mo
- โ Gap Analysis
Phase 4.4: B2B Scale
LATER- โ Batch Generation
- โ "Master Paper" Logic
- โ Classroom Mode
Phase 4.5: Payments
LATER- โ Stripe Integration
- โ Subscription Tiers
- โ Automations
๐ฎFuture Roadmap & Scope
B2B & Institutional Features
- ๐ขSchool/Coaching Center Mode
Support for large-scale user bases. Allowing institutions to generate a single "Master Paper" and distribute it to hundreds of students, while maintaining individual marking and reporting isolation.
- ๐ฅBatch Management
Handling grouped students (e.g., twins, same-year classes) efficiently. One-click generation for multiple profiles with option for "Same Paper" or "Unique Papers" per student.
Business Logic & Monetization
- ๐ณPayment & Subscriptions
Implementation of payment gateways (Stripe/Razorpay) for premium tiers. Tiered access to advanced AI models (GPT-4/Claude-3) and unlimited paper generation.
- ๐Access Control & Limits
- Trial Period: 14-day purely active trial post-registration.
- Student Cap: Max 5 students per parent account.
- Discipline Enforcement: Restrict new paper generation until previous papers for that student are marked/submitted.
๐ปTechnology Stack
Frontend (Cloud Run)
- Next.js 14 (App Router)
- React 18
- TypeScript 5
- Tailwind CSS 3
- Recharts
Backend (Cloud Run)
- FastAPI (Python 3.11)
- SQLAlchemy (Async) / Alembic
- Pydantic v2
- ReportLab (PDF Generation)
- Gmail SMTP (Email Service)
Data & AI
- PostgreSQL 15
- Redis (Planned)
- OpenAI GPT-4o
- Google Gemini Pro
- Anthropic Claude 3.5
DevOps
- Google Cloud Run (Docker)
- GitHub Actions (CI/CD)
- Artifact Registry
- ESLint / Prettier
๐Latest Feature Updates
Restricted Access & Read-Only Mode
- Implemented Read-Only Mode for expired or deactivated accounts.
- Users can log in but cannot generate new papers or download existing ones.
- "Add Student" and "Edit Profile" actions are visually locked for restricted users.
- Smart empty states: "No Students Found" vs "Add Your First Student".
Admin Dashboard Enhancements
- Added Student Email visibility in the Admin Student List.
- Improved student list data fetching to ensure visibility even for deactivated parents.
- Enhanced audit logging for user status changes.
UI/UX Polish
- Subscription Banners: New animated, high-visibility banners for trial and expired states.
- Typography: Refined font weights and spacing in student profiles.
- Accessibility: Improved color contrast and visual hierarchy.
Infrastructure & DevOps
- Automated CI/CD: Full GitHub Actions pipeline deployed to Google Cloud Run.
- Zero-Downtime Deployment: Seamless updates with traffic splitting (supported).
- Cost Optimized: Tuned for maximum efficiency with scale-to-zero.
๐ฐInfrastructure & Engineering Health
Cloud Architecture (GCP)
- Region: europe-west2 (London)
- Project ID: project-cca435f1-a031-459a-ad6
- Compute: Cloud Run (Serverless Container)
- CI/CD: GitHub Actions (Backend: `deploy.yml`, `deploy-dev.yml`)
- Database: Cloud SQL (PostgreSQL) + SQLite (Dev)
- Email: Gmail SMTP (via App Password)
Engineering Health Needs Attention
- Testing: Minimal (Basic Health Checks only). Urgent need for Unit/Integration tests.
- Documentation: Service-level READMEs available. API docs via FastAPI /docs.
- Code Quality: ESLint/Prettier active. Typed Python (Pydantic v2).
Cost Efficiency
- Scale to Zero: Active (Zero idle cost).
- Zero-Trust: Service-to-Service authentication (Planned).