README.md

metadata

title: SmartCertify ML
emoji: 🎓
colorFrom: blue
colorTo: indigo
sdk: docker
pinned: false
license: mit
app_port: 7860

SmartCertify ML Microservice

FastAPI ML service for AI-powered certificate fraud detection.

Upgraded from Render (lightweight) → Hugging Face Spaces (full models):

• Image tampering: ResNet-18 CNN (fine-tuned) — NOT ELA stats
• Similarity: BERT sentence-transformers (all-MiniLM-L6-v2) — NOT TF-IDF
• Chat: DistilBERT zero-shot classification — NOT keyword matching

Endpoints

Method	Path	Description
GET	/health	Health check
POST	/api/ml/verify	Fraud detection — RF+XGB+LGB ensemble
POST	/api/ml/analyze-image	Tampering detection — ResNet-18 CNN
POST	/api/ml/similarity	Duplicate detection — BERT cosine
POST	/api/ml/trust-score	Issuer trust — Gradient Boosting
POST	/api/ml/anomaly	Anomaly detection — Isolation Forest
POST	/api/ml/chat	Q&A chatbot — DistilBERT zero-shot
POST	/api/ml/recommend	Course recs — BERT similarity
GET	/api/ml/metrics	Model metrics

Authentication

All endpoints require X-API-Key header. Set ML_API_KEY as a Space secret in HF settings.

Models

All trained at Docker build time (baked into image):

Model	Type	Size
Fraud detection	RF + XGBoost + LightGBM	~15 MB
Image tampering	ResNet-18 (fine-tuned, CPU)	~45 MB
Semantic similarity	all-MiniLM-L6-v2	~90 MB
Chat classification	DistilBERT zero-shot	~66 MB
Trust scoring	Gradient Boosting	~2 MB
Anomaly detection	Isolation Forest	~1 MB

Local Development

docker build -t smartcertify-ml .
docker run -p 7860:7860 -e ML_API_KEY=dev-key smartcertify-ml
curl http://localhost:7860/health

Deploy to HF Spaces

git remote add hf https://huggingface.co/spaces/YOUR_USERNAME/SmartCertify-ML
git push hf main