live inference api

main.py

@@ -34,7 +34,7 @@ class BatchRequest(BaseModel):
 class BatchResponse(BaseModel):
     inferenceId: str
 
-class InferenceResponse(BaseModel):
+class BatchInferenceResponse(BaseModel):
     inferenceId: str
     status: str
     results: List[Dict]
@@ -55,11 +55,24 @@ async def start_get_recommendations_batch(batch_request: BatchRequest, backgroun
     background_tasks.add_task(process_batch, inferenceId, batch_request.products)
     return BatchResponse(inferenceId=inferenceId)
 
-@app.get("/inference/batch/{inferenceId}", response_model=InferenceResponse)
+@app.get("/inference/batch/{inferenceId}", response_model=BatchInferenceResponse)
 async def get_recommendations_batch(inferenceId: str):
     if inferenceId not in inferences:
         raise HTTPException(status_code=404, detail="Inference ID not found")
     inference = inferences[inferenceId]
-    return InferenceResponse(inferenceId=inferenceId, status=inference["status"], results=inference["result"])
+    return BatchInferenceResponse(inferenceId=inferenceId, status=inference["status"], results=inference["result"])
+
+class InferenceRequest(BaseModel):
+    product: str
+
+class InferenceResponse(BaseModel):
+    cluster: int
+    top_terms: List[str]
 
+# Add a new endpoint for single inferences
+@app.post("/inference", response_model=InferenceResponse)
+def get_recommendations(inference_request: InferenceRequest):
+    cluster_index = show_recommendations(inference_request.product)
+    cluster_terms = get_cluster_terms(cluster_index)
+    return InferenceResponse(cluster=cluster_index, top_terms=cluster_terms)