21

Thanks @Prakul_Agarwal !!

22

Thanks @Akbar_Bakhshi2 !! I understand it know!

23

Hi @Akbar_Bakhshi2 and @Harshad_Dhavale

Thank you so much for your response. I have a couple of question.

Is there any way we can pre-calculate the embedding of documents on mongoDB’s collection?
Suppose my data has 3 fields: title, genres, author. I want to create one more field to store the embedding vector in the collection that stores the original data, then I create a MongoDBAtlasVectorSearch object so I can search directly on it.
(that means I dont have to create a new collection to store text and embedding created by MongoDBAtlasVectorSearch)

24

Hi @Akbar_Bakhshi2 !!! I’m developing a real time RAG approach with Kafka and Atlas Vector, but as Kafka producer is the one in charge of adding the messages to Atlas Vector, I dont know when I should calculate the embeddings for the data inserted in Atlas.
Thanks again!

try:
producer.produce(topic=self.kafka_topic_name, key=str(uuid4()), value=json_string.encode(), on_delivery=self.delivery_report)
producer.flush()
except Exception as ex:
self.get_logger().error(f"Exception: {ex}")