Loyiha 1: Prediction API

🎯 Maqsad

Klassik ML modelni production'a olib chiqadigan to'liq backend servis. Bu loyiha sizning birinchi to'liq portfolio loyihangiz bo'ladi va MLOps'ning asosiy patternlarini ko'rsatadi.

Tavsiya etilgan use case'lar (bittasini tanlang)

Use case	Dataset	Difficulty
Customer Churn Prediction	Telco Customer Churn (Kaggle)	⭐⭐
Loan Default Prediction	LendingClub data	⭐⭐⭐
House Price Estimation	Ames Housing	⭐⭐
Insurance Premium	Kaggle insurance dataset	⭐⭐
Employee Attrition	IBM HR Analytics	⭐⭐⭐
O'zbek dataset	data.gov.uz dataset (extra credit)	⭐⭐⭐⭐

**Maslahat:**Birinchi marta — Churnyoki House Prices.

Architecture

┌─────────────┐      ┌──────────────┐
│  Browser    │─────>│  Streamlit   │
│  Mobile     │      │  Frontend    │
└─────────────┘      └──────┬───────┘
                            │
                            ▼
                     ┌──────────────┐
                     │  FastAPI     │◄─────┐
                     │  Backend     │      │
                     └──────┬───────┘      │
                            │              │
                ┌───────────┼──────────┐   │
                ▼           ▼          ▼   │
        ┌─────────┐  ┌─────────┐  ┌──────────┐
        │ Postgres│  │  Redis  │  │ Sklearn  │
        │ (data)  │  │ (cache) │  │  Model   │
        └─────────┘  └─────────┘  └────┬─────┘
                                       │
                                       ▼
                              ┌──────────────┐
                              │  Prometheus  │
                              │  + Grafana   │
                              └──────────────┘

Tech Stack

Required

**Backend:**FastAPI + Pydantic v2
**ML:**scikit-learn + XGBoost
**Database:**PostgreSQL
**Cache:**Redis
**Container:**Docker + docker-compose
**CI/CD:**GitHub Actions

Nice to have

**Frontend:**Streamlit (oson) yoki React (zo'r)
**Tracking:**MLflow
**Monitoring:**Prometheus + Grafana + Evidently
**Documentation:**mkdocs

Features (must)

MVP (1-hafta)

CSV training pipeline
Sklearn model + serialization
FastAPI /predict endpoint
Pydantic input validation
Docker container
Basic README

V2 (2-hafta)

PostgreSQL — predictions log
Redis caching (same input → cached result)
Batch prediction endpoint
Feedback endpoint (real outcome)
Health checks (liveness, readiness)
Prometheus metrics
Unit + integration tests
GitHub Actions CI

V3 (3-hafta)

MLflow integration (Registry'dan model)
Streamlit dashboard
Drift monitoring (Evidently)
A/B test framework (2 model)
Cloud deployment (Hetzner / Railway / Render)
Blog post
Demo video

API spec

`POST /predict`

// Request
{
    "customer_id": "CUST_12345",
    "tenure_months": 24,
    "monthly_charges": 65.50,
    "total_charges": 1572.00,
    "contract_type": "month-to-month",
    "internet_service": true,
    "payment_method": "credit_card"
}

// Response
{
    "prediction_id": "uuid",
    "customer_id": "CUST_12345",
    "churn_prediction": true,
    "churn_probability": 0.78,
    "risk_level": "high",
    "recommended_action": "send_retention_offer",
    "model_version": "v1.2.3",
    "latency_ms": 23.4
}

`POST /predict/batch`

{
    "customers": [
        {"customer_id": "...", "features": {...}},
        // ...
    ]
}

`POST /feedback`

{
    "prediction_id": "uuid",
    "actual_outcome": true,
    "actual_date": "2026-06-15"
}

`GET /model/info`

{
    "model_name": "churn_predictor",
    "version": "v1.2.3",
    "training_date": "2026-05-15",
    "training_metrics": {
        "accuracy": 0.87,
        "f1": 0.82,
        "auc": 0.91
    },
    "features": [...]
}

`GET /metrics` (Prometheus)

# HELP ml_predictions_total Total predictions
# TYPE ml_predictions_total counter
ml_predictions_total{model_version="v1.2.3",class="0"} 12453
ml_predictions_total{model_version="v1.2.3",class="1"} 3201
...

Project structure

prediction-api/
├── README.md
├── ARCHITECTURE.md
├── docker-compose.yml
├── Dockerfile
├── .env.example
├── .github/
│   └── workflows/
│       ├── ci.yml
│       └── deploy.yml
├── src/
│   ├── api/
│   │   ├── main.py                 # FastAPI app
│   │   ├── routes/
│   │   │   ├── predict.py
│   │   │   ├── feedback.py
│   │   │   └── health.py
│   │   └── schemas.py              # Pydantic models
│   ├── core/
│   │   ├── config.py               # Settings
│   │   └── logging.py
│   ├── data/
│   │   ├── database.py             # SQLAlchemy
│   │   └── models.py               # ORM models
│   ├── ml/
│   │   ├── train.py
│   │   ├── predict.py
│   │   ├── feature_engineering.py
│   │   └── model_registry.py
│   └── monitoring/
│       ├── metrics.py              # Prometheus
│       └── drift.py                # Evidently
├── tests/
│   ├── unit/
│   ├── integration/
│   └── conftest.py
├── notebooks/
│   ├── 01_eda.ipynb
│   ├── 02_feature_engineering.ipynb
│   └── 03_model_training.ipynb
├── data/
│   └── raw/data.csv                # DVC tracked
├── models/
│   └── churn_v1.joblib             # MLflow tracked
├── frontend/                       # Streamlit
│   └── app.py
├── monitoring/
│   ├── prometheus.yml
│   └── grafana-dashboards/
├── pyproject.toml
├── requirements.txt
└── Makefile