use treads instead of processes? Maybe this prevents autokill?

ai_ta_backend/main.py

@@ -1,6 +1,7 @@
 import gc
 import os
 import time
+from concurrent.futures import wait
 from typing import List
 
 from dotenv import load_dotenv
@@ -16,7 +17,8 @@
 app = Flask(__name__)
 CORS(app)
 executor = Executor(app)
-# app.config['EXECUTOR_MAX_WORKERS'] = 5 nothing == picks defaults for me
+app.config['EXECUTOR_TYPE'] = 'thread' # thread is default, or 'process'
+app.config['EXECUTOR_MAX_WORKERS'] = 5
 
 # load API keys from globally-availabe .env file
 load_dotenv()