# US 2026/0093837 A1 · Role-based access control systems for controlling access to a customized knowledge domain for secure ai model development

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

- Number: US 2026/0093837 A1
- Document ID: US20260093837A1
- Status: Published application
- Kind: A1
- Application number: 19/079,529
- Filed: 2025-03-14
- Published: 2026-04-02
- Earliest claimed: 2024-09-30
- Assignees: Truist Bank
- Applicants: Truist Bank
- Inventors: Krishnaveni Kavuri; Chandra Kapireddy; Kenneth William Cluff; Thomas John Mazzaferro; Sasipreetam Morsa
- Web page: https://patentgrep.com/patent/US20260093837A1

## CPC

- G06F 21/6218: {to a system of files or objects, e.g. local or distributed file system or database}
- H04L 63/105: {Multiple levels of security}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Systems and methods provide an operative connection to existing large language model(s) to be included in the customized knowledge domain framework. Further, instructions on use and deployment to be included as selectable metadata for establishing guardrails for a generative artificial intelligence application are obtained, where the guardrails establish rules for integration and use of the generative artificial intelligence application. Parser(s) for selection to be applied to the generative artificial intelligence application are established. Selection of prompt template(s) for structuring user inputs and model outputs for the generative artificial intelligence application is facilitated, and an operative connection to existing agentic large language models for selection is provided. Further, access to existing validation large language model(s) is provided for selection.

## Claims

### Claim 1

A computing system for role-based access control to regulate access to AI models, the system comprising:
at least one processor;
a communication interface communicatively coupled to the at least one processor; and
a memory device storing executable code that, when executed, causes the at least one processor to:
provide an operative connection to one or more existing large language models to be included in a customized knowledge domain framework, wherein access to portions of the customized knowledge domain framework are restricted in accordance with role-based access controls (RBAC);
obtain further instructions associated with acceptable use and deployment to be included as selectable metadata for establishing guardrails for a generative artificial intelligence application that incorporates information from the customized knowledge domain framework, wherein the guardrails establish rules for integration and use of the generative artificial intelligence application;
establish one or more parsers for selection to be applied to the generative artificial intelligence application;
facilitate selection of one or more prompt templates for structuring user inputs and model outputs for the generative artificial intelligence application;
provide an operative connection to one or more existing agentic large language models for selection; and
provide access to one or more existing validation large language models for selection.

### Claim 2 (depends on 1)

The computing system of claim 1, wherein the one or more existing large language models are selected from a curated list of existing large language models, the one or more existing large language models being supported with one or more secure application programming interfaces.

### Claim 3 (depends on 1)

The computing system of claim 1, wherein the rules include at least one selected from the group consisting of relevance, coherence, an Ada similarity score, and integration capabilities in relation to an entity's governance platforms.

### Claim 4 (depends on 1)

The computing system of claim 1, wherein the one or more parsers include input parsers for parsing unstructured content for inclusion in the generative artificial intelligence application.

### Claim 5 (depends on 4)

The computing system of claim 4, wherein the input parsers are configured to parse one or more files selected from the group consisting of HTML files and PDF files.

### Claim 6 (depends on 4)

The computing system of claim 4, wherein the input parsers are selected from the group consisting of an optical character recognition (OCR) parser and a layout parser.

### Claim 7 (depends on 1)

The computing system of claim 1, wherein the one or more parsers include output parsers, the output parsers being configured to parse an output from the generative artificial intelligence application and parse the output into a structured format.

### Claim 8 (depends on 1)

The computing system of claim 1, wherein the executable code, when executed, further causes the at least one processor to receive an input indicating a corpus of documents to be included within the customized knowledge domain framework of an entity and based thereon apply the corpus of documents to one or more custom neural document models that combine language and layout features to extract labeled fields from the corpus of documents, the one or more custom neural document models being trained on a plurality of document types such that the labeled fields are extractable from structured, unstructured, and linked data products.

### Claim 9 (depends on 8)

The computing system of claim 8, wherein the executable code, when executed, further causes the at least one processor to train the one or more custom neural document models, the training including:
inserting the training data into an iterative training and testing loop to predict a target variable; and
repeatedly predicting the target variable during each iteration of the training and testing loop, wherein each iteration of the training and testing loop has differing weights applied to one or more nodes of the one or more custom neural document models, each of the differing weights being updated with each iteration of the training and testing loop to reduce error in predicting the target variable, which improves predictability of the target variable and functionality of the one or more custom neural document models.

### Claim 10 (depends on 1)

The computing system of claim 1, wherein the generative artificial intelligence application incorporates information from the one or more existing large language models, the one or more existing agentic large language models, and the one or more existing validation large language models.

### Claim 11 (depends on 1)

The computing system of claim 1, wherein the generative artificial intelligence application incorporates information from at least one of (a) the one or more existing large language models, (b) the one or more existing agentic large language models, and (c) the one or more existing validation large language models.

### Claim 12 (depends on 1)

The computing system of claim 1, wherein the generative artificial intelligence application incorporates information from at least two of (a) the one or more existing large language models, (b) the one or more existing agentic large language models, and (c) the one or more existing validation large language models.

### Claim 13 (depends on 1)

The computing system of claim 1, wherein the executable code, when executed, further causes the at least one processor to initiate creation of the generative artificial intelligence application based on receiving, via the one or more prompt templates, an indication of the information to be included from the customized knowledge domain framework.

### Claim 14 (depends on 1)

The computing system of claim 1, wherein the customized knowledge domain framework is exclusive to data sources internal to an enterprise.

### Claim 15 (depends on 14)

The computing system of claim 14, wherein the RBAC restricts network access to the customized knowledge domain framework based on a role of an individual within the enterprise, the RBAC defining access to an underlying corpus of documents included in the portions of the customized knowledge domain framework.

### Claim 16 (depends on 1)

The computing system of claim 1, wherein the one or more parsers are created in response to one or more definition inputs.

### Claim 17 (depends on 1)

The computing system of claim 1, wherein the instructions associated with the acceptable use and deployment align with an existing guardrail, and based thereon the executable code, when executed, further causes the at least one processor to include the existing guardrail for the guardrails for the generative artificial intelligence application.

### Claim 18 (depends on 1)

The computing system of claim 1, wherein the customized knowledge domain framework is programmed with default settings for generating the generative artificial intelligence application, the default settings including embedded models, retrievers, rankers, and a retrieval-augmented generation score.

### Claim 19

A non-transitory computer-readable storage medium the computer-readable storage medium including instructions that when executed by a processor, cause the processor to:
provide an operative connection to one or more existing large language models to be included in a customized knowledge domain framework, wherein access to portions of the customized knowledge domain framework are restricted in accordance with role-based access controls (RBAC);
obtain further instructions associated with acceptable use and deployment to be included as selectable metadata for establishing guardrails for a generative artificial intelligence application that incorporates information from the customized knowledge domain framework, wherein the guardrails establish rules for integration and use of the generative artificial intelligence application;
establish one or more parsers for selection to be applied to the generative artificial intelligence application;
facilitate selection of one or more prompt templates for structuring user inputs and model outputs for the generative artificial intelligence application;
provide an operative connection to one or more existing agentic large language models for selection; and
provide access to one or more existing validation large language models for selection.

### Claim 20

A computer-implemented method, comprising:
providing an operative connection to one or more existing large language models to be included in a customized knowledge domain framework, wherein access to portions of the customized knowledge domain framework are restricted in accordance with role-based access controls (RBAC);
obtaining instructions associated with acceptable use and deployment to be included as selectable metadata for establishing guardrails for a generative artificial intelligence application that incorporates information from the customized knowledge domain framework, wherein the guardrails establish rules for integration and use of the generative artificial intelligence application;
establishing one or more parsers for selection to be applied to the generative artificial intelligence application;
facilitating selection of one or more prompt templates for structuring user inputs and model outputs for the generative artificial intelligence application;
providing an operative connection to one or more existing agentic large language models for selection; and
providing access to one or more existing validation large language models for selection.
