Providing personalized prompts to users based on documents in cloud storage
Systems and methods include pre-processing documents in cloud storage using query embeddings, providing personalized prompts to users based on documents in cloud storage, real-time anticipation of user interest in information contained in documents in cloud storage, and providing generative answers including citation to source documents in cloud storage. The system and methods generate generative machine learning model (MLM) prompts based on document portions of documents in a cloud-based content management platform. The systems and methods use the generative MLM to generate responses to prompts, and the responses include citations to the document portions used to generate the responses in order for users to verify the responses.
1 . A method, comprising:
selecting, based on a selection criterion, a plurality of documents stored in a cloud-based content management platform;
selecting a portion of a document in the plurality of documents;
generating, using a generative machine learning model (MLM), a first generative MLM prompt based on the portion of the document;
associating the first generative MLM prompt with the portion of the document; and
causing the first generative MLM prompt to be presented to a user of the cloud-based content management platform when the user accesses the cloud-based content management platform.
2 . The method of claim 1 , wherein selecting the plurality of documents based on the selection criterion comprises selecting one or more documents whose metadata indicate that the one or more documents were last modified within a predetermined time threshold.
3 . The method of claim 1 , wherein selecting the plurality of documents based on the selection criterion comprises selecting one or more documents whose metadata indicate that the one or more documents were last opened within a predetermined time threshold.
4 . The method of claim 1 , wherein selecting the plurality of documents based on the selection criterion comprises selecting one or more documents whose metadata indicate that the one or more documents were generated within a predetermined time threshold.
5 . The method of claim 1 , wherein selecting the plurality of documents based on the selection criterion comprises selecting one or more documents whose metadata indicate that the one or more documents each received a comment from a user of the cloud-based content management platform within a predetermined time threshold.
6 . The method of claim 1 , wherein selecting the portion of the document comprises:
retrieving a query embedding associated with the document;
generating, via an embedding model, a query embedding based on the portion of the document; and
responsive to the query embedding of the portion of the document being within a threshold similarity from the retrieved query embedding associated with the document, selecting the portion of the document.
7 . The method of claim 6 , wherein the threshold similarity comprises a cosine similarity.
8 . The method of claim 1 , wherein the first generative MLM prompt is further based on a second generative MLM prompt comprising:
a command for the generative MLM to generate the first generative MLM prompt; and
the portion of the document.
9 . The method of claim 1 , wherein causing the first generative MLM prompt to be presented to the one or more users comprises providing, on a user interface of the cloud-based content management platform, a selectable option to input the first generative MLM prompt into a text field.
10 . The method of claim 9 , wherein providing the selectable option comprises providing the selectable option below a search field of the user interface.
11 . A system, comprising:
a memory; and
one or more processing devices, coupled to the memory, configured to perform operations comprising:
selecting, based on a selection criterion, a plurality of documents stored in a cloud-based content management platform;
selecting a portion of a document of the plurality of documents;
generating, using a generative machine learning model (MLM), a first generative MLM prompt based on:
the portion of the document, and
a second generative MLM prompt, wherein the second generative MLM prompt comprises a command for the generative MLM to generate a generative MLM prompt about the portion of the document;
associating the first generative MLM prompt with the document; and
causing the first generative MLM prompt to be presented to one or more users of the cloud-based content management platform when the one or more users access the cloud-based content management platform.
12 . The system of claim 11 , wherein the generative MLM comprises a transformer-based large language model.
13 . The system of claim 11 , wherein the command to generate the first generative MLM prompt about the portion of the document comprises a command to generate a generative MLM prompt to summarize the plurality of documents.
14 . The system of claim 11 , wherein the command to generate the generative MLM prompt about the portion of the document comprises a command to generate a generative MLM prompt to identify modifications to the plurality of documents.
15 . The system of claim 11 , wherein the selection criterion comprises a query embedding of each document in the plurality of documents being within a threshold similarity from a query embedding associated with a document that the one or more users opened within a threshold amount of time.
16 . The system of claim 11 , wherein selecting the portion of the document comprises selecting the entire document.
17 . A non-transitory, computer-readable storage medium storing instructions that, when executed, cause a processing device to:
select, based on a selection criterion, one or more documents stored in a cloud-based content management platform;
select a portion of a document of the one or more documents;
generate, using a generative machine learning model (MLM), a generative MLM prompt based on the portion of the document;
associate the generative MLM prompt with the document; and
cause the generative MLM prompt to be presented to one or more users of the cloud-based content management platform when the one or more users access the cloud-based content management platform.
18 . The computer-readable storage medium of claim 17 , wherein the one or more documents comprises one or more documents stored in the same folder in the cloud-based content management platform.
19 . The computer-readable storage medium of claim 17 , wherein:
each document in the one or more documents comprises metadata, wherein the metadata includes data indicating an owner user of the respective document; and
the metadata of the one or more documents comprises data indicating that at least two documents of the one or more documents have different owner users.
20 . The computer-readable storage medium of claim 17 , wherein:
the document comprises a plurality of pages; and
selecting the portion of the document comprises selecting a predetermined number of pages of the plurality of pages.