# US 2026/0064896 A1 · Artificial Intelligence (AI) Assisted Digital Documentation for Digital Engineering

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

- Number: US 2026/0064896 A1
- Document ID: US20260064896A1
- Status: Published application
- Kind: A1
- Application number: 19/381,004
- Filed: 2025-11-06
- Published: 2026-03-05
- Earliest claimed: 2023-02-01
- Assignees: Istari Digital, Inc.
- Applicants: Istari Digital, Inc.
- Inventors: William Roper, JR.; Christopher Lee Benson; Sriram Krishnan; Peter Galvin; Pranav Sumanth Doijode; Bayan Abedal-Muttaleb Afif Hashem
- Web page: https://patentgrep.com/patent/US20260064896A1

## CPC

- G06F 30/10: Geometric CAD
- G06F 40/186: Templates
- G06F 30/15: Vehicle, aircraft or watercraft design
- G06F 30/27: using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
- G06F 40/10: Text processing (natural language analysis G06F40/20; semantic analysis G06F40/30; processing or translation of natural language G06F40/40)
- G06F 40/30: Semantic analysis
- G06N 3/006: based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0455: Auto-encoder networks; Encoder-decoder networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/047: Probabilistic or stochastic networks
- G06N 3/0475: Generative networks
- G06N 3/048: Activation functions
- G06N 3/08: Learning methods
- G06N 3/088: Non-supervised learning, e.g. competitive learning
- G06N 3/092: Reinforcement learning
- G06N 3/096: Transfer learning
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06N 5/02: Knowledge representation; Symbolic representation
- G06N 5/045: Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence
- G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- G06N 20/20: Ensemble learning
- G06F 2111/02: CAD in a network environment, e.g. collaborative CAD or distributed simulation
- G06F 2111/20: Configuration CAD, e.g. designing by assembling or positioning modules selected from libraries of predesigned modules
- G06N 3/126: Evolutionary algorithms, e.g. genetic algorithms or genetic programming
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A digital documentation system for preparation of engineering documents utilizing one or more artificial intelligence (AI) algorithms is provided. The system includes a user interface for selecting and populating templates with data, and one or more AI algorithms for creating and recommending templates, and preparing documents based on the recommended templates. The system uses natural language processing and semantic analysis algorithms to understand the content of the templates, documents, and associated engineering data, and to generate and recommend relevant templates to the user based on user prompts. The system also uses machine learning and predictive modeling and decision-tree algorithms to assist with the preparation of documents, by generating suggestions for data fields and values based on the user's previous inputs and the overall context of the document and available engineering data, including model data and metadata from digital models accessed in a zero-trust framework.

## Claims

### Claim 1

A non-transitory physical storage medium storing instructions, the instructions executable by a processor of a digital documentation system to cause the processor to perform operations comprising:
retrieve one or more digital engineering (DE) document templates from a DE document template library comprising DE document templates for one or more phases of a DE product lifecycle, wherein the DE document templates comprise first DE data fields;
receive a first user input from a user, wherein the first user input comprises a model splice of a first DE model file of a DE model type, wherein the first DE model file resides in a customer environment distinct from the digital documentation system, wherein the model splice was generated by wrapping the first DE model file with an externally-accessible API associated with the DE model type, wherein the model splice provides access to limited portions of a plurality of model data of the first DE model file, and wherein the model splice provides access control to the plurality of model data based on an access permission of the user;
select a DE document template from the one or more DE document templates based on the first user input;
generate a first model data using a first splice function of the model splice of the first DE model file, via the externally-accessible API, based on the selected DE document template;
train a generator machine learning (ML) model on one or more example DE document files and a platform documentation from the digital documentation system to generate a trained generator ML model, wherein the platform documentation comprises a reference guide of the externally-accessible API; and
execute the trained generator ML model to generate a DE document file from the selected DE document template and the first model data.

### Claim 2 (depends on 1)

The non-transitory physical storage medium of claim 1, wherein the instructions to generate the model splice further comprise instructions to:
crawl the first DE model file of the DE model type to extract the plurality of model data from the first DE model file to generate a data schema describing a structure and a format of the plurality of model data.

### Claim 3 (depends on 2)

The non-transitory physical storage medium of claim 2, wherein the instructions to generate the model splice further comprise instructions to:
generate a plurality of splice function scripts to access the plurality of model data based on the access permission of the user,
wherein the plurality of splice function scripts are written in a computer-executable scripting language,
wherein at least one of the plurality of splice function scripts invokes a native API function of a third-party DE tool, and
wherein the plurality of splice function scripts provide a plurality of API endpoints to access and manipulate the plurality of model data based on the data schema.

### Claim 4 (depends on 3)

The non-transitory physical storage medium of claim 3, wherein the instructions to generate the model splice further comprise instructions to:
generate the externally-accessible API from the plurality of splice function scripts,
wherein the externally-accessible API allows access to the plurality of model data without directly invoking the native API function of the third-party DE tool.

### Claim 5 (depends on 1)

The non-transitory physical storage medium of claim 1, further comprising instructions to:
fine-tune the generator ML model using a fine-tuning database to generate a fine-tuned generator ML model,
wherein the fine-tuning database comprises at least a sample prompt-response pair, and
wherein the sample prompt-response pair comprises a sample user input and a corresponding response of the digital documentation system.

### Claim 6 (depends on 1)

The non-transitory physical storage medium of claim 1, wherein the generation of the first model data is executed upon invocation by an orchestration script of the externally-accessible API.

### Claim 7 (depends on 1)

The non-transitory physical storage medium of claim 1, further comprising instructions to:
receive one or more modifications to the first model data upon an update of the first DE model file via the model splice; and
update one or more portions of the DE document file based on the one or more modifications to the first model data.

### Claim 8 (depends on 7)

The non-transitory physical storage medium of claim 7, further comprising instructions to:
detect a first modification of the first DE model file or a second modification of a software-defined digital thread associated with the first DE model file;
update the first DE data fields of the DE document file based on the first modification or the second modification responsive to the first modification or the second modification; and
generate an updated printed DE document file with the updated first DE data fields updated based on the DE document file.

### Claim 9 (depends on 1)

The non-transitory physical storage medium of claim 1, further comprising instructions to:
receive user feedback data related to the DE document file generated by the generator ML model from the user;
generate feedback metrics related to a quality of the DE document file generated by the generator ML model; and
train and/or fine-tune the generator ML model utilizing the feedback metrics to improve future DE document files generated by the generator ML model.

### Claim 10 (depends on 1)

The non-transitory physical storage medium of claim 1, further comprising instructions to:
generate training data comprising a plurality of DE document files from the generator ML model and document edits made to the plurality of DE document files by the user; and
train and/or fine-tune the generator ML model on the training data.
