feat(ai-services): add Python FastAPI AI/ML services container

Create libs/ai-services/ with FastAPI app providing:
- POST /avm/predict — XGBoost-backed property price prediction (heuristic fallback)
- POST /avm/extract-features — Vietnamese NLP feature extraction from listing text
- POST /moderation/check — content moderation with rule-based flagging
- GET /health — health check endpoint

Includes Dockerfile (Python 3.12), docker-compose integration, Pydantic models,
and 9 passing tests covering all endpoints.

Co-Authored-By: Paperclip <noreply@paperclip.ing>

libs/ai-services/tests/test_avm.py

@@ -0,0 +1,59 @@
+from fastapi.testclient import TestClient
+
+from app.main import app
+
+client = TestClient(app)
+
+
+def test_predict_heuristic():
+    resp = client.post(
+        "/avm/predict",
+        json={
+            "area": 80.0,
+            "district": "Cầu Giấy",
+            "city": "Hà Nội",
+            "property_type": "apartment",
+            "bedrooms": 2,
+            "bathrooms": 2,
+            "floors": 1,
+            "frontage": 0,
+            "road_width": 0,
+            "has_legal_paper": True,
+        },
+    )
+    assert resp.status_code == 200
+    data = resp.json()
+    assert data["estimated_price_vnd"] > 0
+    assert 0 <= data["confidence"] <= 1
+    assert data["price_per_m2"] > 0
+    assert data["price_range_low"] < data["estimated_price_vnd"]
+    assert data["price_range_high"] > data["estimated_price_vnd"]
+
+
+def test_predict_validation_error():
+    resp = client.post(
+        "/avm/predict",
+        json={"area": -10, "district": "", "city": "HN", "property_type": "house"},
+    )
+    assert resp.status_code == 422
+
+
+def test_extract_features():
+    text = "Bán căn hộ chung cư 80m2 3 phòng ngủ 2 WC tầng 10 giá 3.5 tỷ sổ đỏ chính chủ"
+    resp = client.post("/avm/extract-features", json={"text": text})
+    assert resp.status_code == 200
+    data = resp.json()
+    features = data["features"]
+    assert features["area"] == 80.0
+    assert features["bedrooms"] == 3
+    assert features["bathrooms"] == 2
+    assert features["property_type"] == "apartment"
+    assert features["has_legal_paper"] is True
+    assert features["price_mentioned"] == 3_500_000_000
+
+
+def test_extract_features_minimal():
+    resp = client.post("/avm/extract-features", json={"text": "Bán nhà riêng"})
+    assert resp.status_code == 200
+    data = resp.json()
+    assert data["features"]["property_type"] == "house"

libs/ai-services/tests/test_health.py

@@ -0,0 +1,12 @@
+from fastapi.testclient import TestClient
+
+from app.main import app
+
+client = TestClient(app)
+
+
+def test_health():
+    resp = client.get("/health")
+    assert resp.status_code == 200
+    data = resp.json()
+    assert data["status"] == "ok"

libs/ai-services/tests/test_moderation.py

@@ -0,0 +1,50 @@
+from fastapi.testclient import TestClient
+
+from app.main import app
+
+client = TestClient(app)
+
+
+def test_clean_text():
+    resp = client.post(
+        "/moderation/check",
+        json={"text": "Bán căn hộ đẹp tại quận 1", "context": "listing"},
+    )
+    assert resp.status_code == 200
+    data = resp.json()
+    assert data["is_flagged"] is False
+    assert data["score"] == 0.0
+
+
+def test_phone_number_flagged():
+    resp = client.post(
+        "/moderation/check",
+        json={"text": "Liên hệ 0912345678 để xem nhà", "context": "listing"},
+    )
+    assert resp.status_code == 200
+    data = resp.json()
+    assert data["is_flagged"] is True
+    assert any(f["category"] == "contact_info" for f in data["flags"])
+    assert "[REDACTED]" in data["cleaned_text"]
+
+
+def test_scam_language_flagged():
+    resp = client.post(
+        "/moderation/check",
+        json={"text": "Cảnh báo lừa đảo từ chủ nhà", "context": "comment"},
+    )
+    assert resp.status_code == 200
+    data = resp.json()
+    assert data["is_flagged"] is True
+    assert any(f["category"] == "profanity" for f in data["flags"])
+
+
+def test_prohibited_property():
+    resp = client.post(
+        "/moderation/check",
+        json={"text": "Bán lô đất rừng phòng hộ 500m2", "context": "listing"},
+    )
+    assert resp.status_code == 200
+    data = resp.json()
+    assert data["is_flagged"] is True
+    assert any(f["category"] == "prohibited_content" for f in data["flags"])