# US 2025/0211549 A1 · Context-based generative artificial intelligence system

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0211549 A1
- Document ID: US20250211549A1
- Status: Published application
- Kind: A1
- Application number: 18/390,754
- Filed: 2023-12-20
- Published: 2025-06-26
- Earliest claimed: 2023-12-20
- Assignees: Paypal, Inc.
- Applicants: PayPal, Inc.
- Inventors: Elavarasi Arunachalam; Srinivasa Sai Kaushal Bondada
- Web page: https://patentgrep.com/patent/US20250211549A1

## CPC

- H04L 51/02: using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Systems, methods, and computer program products for using a generative artificial intelligence system to generate answers or summaries is provided. During the ingestion stage, the system receives documents and transcripts that include data associated with a theme. The data is converted into a common format and is divided into chunks. The chunks are associated with metadata tags that include chunk and data information. From the chunks, the system generates embedding vectors. During the inference stage, the system receives an information request. If the information request is a question, the system generates a vector from the question, and uses a similarity search to identify similar vectors. From the similar vectors, the system identifies chunks. If the information request includes a summary request, the system uses the metadata tags to identify chunks with summary information. The system generates an answer or a summary from the identified chunks.

## Claims

### Claim 1

A method comprising:
generating a project identifier associated with project information;
generating a plurality of chunks from the project information and a plurality of metadata tags, wherein each chunk in the plurality of chunks corresponds to a respective metadata tag in the plurality of metadata tags;
generating, using an embedding large language model, a plurality of vectors from the plurality of chunks, wherein the each chunk corresponds to a respective vector in the plurality of vectors;
generating a project dictionary associated with the project identifier, the plurality of chunks, the plurality of metadata tags, and the plurality of vectors; and
storing the project dictionary in a vector database with project dictionaries associated with other projects.

### Claim 2 (depends on 1)

The method of claim 1, wherein the project information comprises at least one document and at least one transcript.

### Claim 3 (depends on 2)

The method of claim 2, wherein generating the plurality of chunks further comprises:
identifying sections of text in the at least one document or the at least one transcript;
dividing the sections of text into a subset of chunks in the plurality of chunks using at least one rule; and
associating a subset of tags in the plurality of tags with the subset of chunks, wherein the subset of tags comprises an identifier associated with the at least one document or the at least one transcript.

### Claim 4 (depends on 1)

The method of claim 1, wherein the project information comprises data from multiple sources.

### Claim 5 (depends on 1)

The method of claim 1, wherein the respective metadata tag includes at least the project identifier associated with the project information.

### Claim 6 (depends on 1)

The method of claim 1, wherein generating the plurality of chunks further comprises:
recursively dividing sections of text in the at least one document or the at least one transcript to generate the plurality of chunks until the each chunk in the plurality of chunks is less than a predefined chunk size.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
receiving the project identifier and a request for information at a chatbot interface associated with a generative artificial intelligence language model;
selecting the project dictionary using the project identifier;
determining that the request for the information is a question;
converting, using the generative artificial intelligence language model, the question into a vector;
retrieving, using a similarity algorithm, a subset of vectors from the plurality of vectors associated with the project dictionary that the similarity algorithm determined to be similar to the vector; and
converting, using the generative artificial intelligence language model, a subset of chunks associated with the subset of vectors into an answer to the question.

### Claim 8 (depends on 1)

The method of claim 1, further comprising:
receiving the project identifier and a request for information from a chatbot interface associated with a generative artificial intelligence language model;
selecting the project dictionary using the project identifier;
determining that the request for the information is a request for a summary;
retrieving, using the plurality of metadata tags associated with the project dictionary, a subset of chunks in the plurality of chunks; and
generating the summary using the subset of chunks in the plurality of chunks, wherein the summary is transmitted for display at the chatbot interface.

### Claim 9

A system comprising:
one or more non-transitory memories storing instructions; and
one or more hardware processors coupled to the non-transitory memories and configured to read the instructions from the non-transitory memories to cause the system to perform operations comprising:
receiving an identifier and a request for information from a chatbot interface;
generating, using at least one large language model, a vector from the request for the information;
accessing a dictionary from a plurality of dictionaries with the identifier, the dictionary including a plurality of vectors associated with a plurality of chunks, wherein the plurality of chunks corresponds to data associated with the identifier from a plurality of data sources;
identifying, using a similarity algorithm, a subset of vectors from the plurality of vectors in the dictionary that are similar to the vector associated with the request for the information, wherein the subset of vectors that are similar comprise parameters within a predefined distance from parameters of the vector associated with the request for the information;
identifying a subset of chunks corresponding to the subset of vectors; and
generating, using the at least one large language model, a response to the request for the information from the subset of chunks.

### Claim 10 (depends on 9)

The system of claim 9, wherein the request for the information is included in a dialogue between an artificial intelligence chatbot and a user.

### Claim 11 (depends on 9)

The system of claim 9, further comprising:
ranking the subset of vectors according to similarity indicated by the similarity algorithm from most similar to least similar; and
retaining a predefined number of most vectors in the subset of vectors.

### Claim 12 (depends on 11)

The system of claim 11, wherein identifying the subset of chunks further comprises:
identifying the subset of chunks in an order of the ranked subset of vectors; and
generating the response to the request using the subset of chunks in the order of the ranked subset of vectors.

### Claim 13 (depends on 9)

The system of claim 9, wherein the request for the information is associated with a software project.

### Claim 14 (depends on 9)

The system of claim 9, wherein the information is from a plurality of sources and includes at least one document or at least one transcript that include alphanumeric text.

### Claim 15 (depends on 9)

The system of claim 9, wherein the at least one large language model is finetuned using historical data from a plurality of software projects.

### Claim 16 (depends on 9)

The system of claim 9, further comprising:
determining that the request for the information is a question associated with a project, and wherein the response is an answer to the question.

### Claim 17

A non-transitory machine-readable medium having stored thereon machine-readable instructions executable to cause a machine to perform operations comprising:
receiving an identifier and a request for information from a chatbot interface;
accessing a dictionary from a plurality of dictionaries with the identifier, the dictionary including a plurality of chunks, wherein the plurality of chunks corresponds to data associated with a plurality of data sources;
retrieving at least one chunk from the plurality of chunks associated with the dictionary; and
generating a response to the request for the information using the at least one chunk.

### Claim 18 (depends on 17)

The non-transitory machine-readable medium of claim 17, wherein the request for the information is the request for a summary of a project.

### Claim 19 (depends on 17)

The non-transitory machine-readable medium of claim 17, wherein the dictionary further comprises a plurality of metadata tags, wherein a subset of metadata tags identify a title associated with a data source in the plurality of data sources.

### Claim 20 (depends on 19)

The non-transitory machine-readable medium of claim 19, wherein the retrieving the at least one chunk further comprises:
identifying the at least one chunk corresponding to the subset of metadata tags.
