Quantara AI is a web platform that automates the creation of engineering laboratory reports in realistic handwritten notebook format.
Students upload their experiment details, readings, and handwriting samples. The platform generates a complete lab report, renders it in the student's handwriting style using blue ink, draws hand-sketched diagrams and graphs, and exports the result as a notebook-style PDF.
The goal is to eliminate the manual effort of writing lengthy lab reports while preserving the handwritten format required by many institutions.
- Register/Login
- Dashboard
- Slot-based submission system
- Request history
- Download generated reports
Students provide:
- Course code
- Experiment title
- Aim
- Apparatus
- Theory notes (optional)
- Raw observations/readings
- Lecturer instructions
- Preferred report format
AI generates:
- Title page
- Aim
- Apparatus
- Theory
- Procedure
- Observation tables
- Calculations
- Results
- Discussion
- Conclusion
- References
Students upload:
- 5–10 pages of handwritten notes
System extracts:
- Character shapes
- Letter spacing
- Writing slant
- Writing size
System creates a reusable handwriting profile.
Generated report is converted into:
- Blue pen handwriting
- Notebook paper pages
- Natural spacing variations
- Line-by-line writing
- Multi-page report format
Automatically generates:
- Circuit diagrams
- Block diagrams
- Engineering apparatus sketches
- Graphs
- Tables
- Flowcharts
Rendered to look hand drawn rather than computer generated.
Output:
- Multi-page notebook PDF
- Blue ink handwriting
- Hand-drawn diagrams
- Ready for printing or copying
Frontend (React + Vite)
↓
Express API
↓
PostgreSQL
↓
BullMQ + Redis Queue
↓
Python AI Service (FastAPI)
↓
Report Generator + Handwriting Engine + Diagram Renderer
- React
- Vite
- TypeScript
- TailwindCSS
- TanStack Query
- React Router
- shadcn/ui
- Node.js
- Express.js
- Prisma ORM
- PostgreSQL
- JWT Authentication
- Multer
- BullMQ
- Redis
- Python
- FastAPI
- PyTorch
- Pillow
- OpenCV
- NumPy
OpenRouter
Benefits:
- Free models available
- OpenAI-compatible API
- Can switch models without changing code
Recommended Models:
- DeepSeek
- Qwen
- Gemma
- Llama
Expected Cost:
- Free during development
- Under ₦5,000/month for early usage if using inexpensive models
Font-based approach
Tools:
- FontForge
- Calligraphr
- TraceFont
Benefits:
- Free or low cost
- Fast implementation
- Good enough for MVP
AI handwriting synthesis
Research Projects:
- GANWriting
- DeepWriting
- Diffusion-based handwriting models
Benefits:
- More realistic output
- Can imitate individual writing styles
Drawback:
- Higher complexity
Purpose:
Creates diagrams that appear hand drawn.
Used for:
- Graphs
- Circuits
- Flowcharts
- Engineering sketches
Cost:
Free and open source
Tools:
- Pillow
- OpenCV
Responsibilities:
- Blue pen effect
- Paper texture
- Notebook lines
- Ink variations
- Page composition
Cost:
Free
Options:
- pdf-lib
- PDFKit
Cost:
Free
Tools:
- Redis
- BullMQ
Workflow:
Student submits report
↓
Queue job
↓
Generate report
↓
Generate handwriting
↓
Generate diagrams
↓
Generate notebook pages
↓
Generate PDF
↓
Notify student
users
handwriting_profiles
lab_reports
generated_pages
slots
payments
notifications
Development Stage:
₦0 – ₦5,000/month
Production (First 100–300 Users):
₦5,000 – ₦20,000/month
Major expenses:
- Hosting
- Database
- AI API usage
Most rendering, handwriting generation, PDF creation, and diagram generation can remain completely open source.
A student uploads experiment details and handwriting samples.
The system automatically:
- Generates a complete lab report.
- Renders it in blue handwritten style.
- Generates hand-drawn diagrams and graphs.
- Places everything on notebook paper.
- Exports a multi-page PDF.
Result:
A realistic handwritten engineering lab report generated with minimal human effort.