PubMed Download Pipeline

ai_ta_backend/main.py

@@ -40,6 +40,7 @@
 from ai_ta_backend.service.sentry_service import SentryService
 
 from ai_ta_backend.beam.nomic_logging import create_document_map
+from ai_ta_backend.utils.pubmed_extraction import extractPubmedData
 
 app = Flask(__name__)
 CORS(app)
@@ -379,6 +380,17 @@ def getTopContextsWithMQR(service: RetrievalService, posthog_service: PosthogSer
   response.headers.add('Access-Control-Allow-Origin', '*')
   return response
 
+@app.route('/pubmedExtraction', methods=['GET'])
+def pubmedExtraction():
+  """
+  Extracts metadata and download papers from PubMed.
+  """
+  result = extractPubmedData()
+
+  response = jsonify(result)
+  response.headers.add('Access-Control-Allow-Origin', '*')
+  return response
+
 
 def configure(binder: Binder) -> None:
   binder.bind(RetrievalService, to=RetrievalService, scope=RequestScope)