WIP: moving to services, refining matrix_service

ai_service/Dockerfile

@@ -0,0 +1,12 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+COPY requirements.txt ./
+
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY main.py ./
+
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000"]
+

ai_service/main.py

@@ -0,0 +1,45 @@
+import os
+from dotenv import load_dotenv
+from fastapi import FastAPI, Header, HTTPException
+from pydantic import BaseModel
+import openai
+import redis
+
+AI_TOKEN = os.environ["AI_HANDLER_TOKEN"]
+openai.api_key = os.environ["OPENAI_API_KEY"]
+r = redis.Redis.from_url(os.environ.get("REDIS_URL", "redis://redis:6379"))
+
+class MessagePayload(BaseModel):
+    roomId: str
+    userId: str
+    content: str
+    eventId: str
+    timestamp: int
+
+app = FastAPI()
+
+@app.post("/api/v1/message")
+async def message(
+    payload: MessagePayload,
+    authorization: str = Header(None)
+):
+    if authorization != f"Bearer {AI_TOKEN}":
+        raise HTTPException(status_code=401, detail="Unauthorized")
+
+    # Idempotency: ignore duplicates
+    if r.get(payload.eventId):
+        return {"reply": r.get(payload.eventId).decode()}
+
+    # Build prompt (very simple example)
+    prompt = f"User {payload.userId} said: {payload.content}\nBot:"
+    resp = openai.Completion.create(
+        model="text-davinci-003",
+        prompt=prompt,
+        max_tokens=150
+    )
+    reply = resp.choices[0].text.strip()
+
+    # Cache reply for idempotency
+    r.set(payload.eventId, reply, ex=3600)
+
+    return {"reply": reply}

ai_service/requirements.txt

@@ -0,0 +1,6 @@
+python-dotenv>=1.0.0
+openai
+fastapi>=0.95
+uvicorn>=0.22
+redis>=4.5
+pydantic>=1.10