# US 2025/0053861 A1 · Compute agnostic machine learning project workspaces

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

- Number: US 2025/0053861 A1
- Document ID: US20250053861A1
- Status: Published application
- Kind: A1
- Application number: 18/447,615
- Filed: 2023-08-10
- Published: 2025-02-13
- Earliest claimed: 2023-08-10
- Assignees: Optum, Inc.
- Applicants: Optum, Inc.
- Inventors: Cory MUIR; Anand DHANDHANIA; Vivek BHADAURIA; Vasant MANOHAR
- Web page: https://patentgrep.com/patent/US20250053861A1

## CPC

- G06N 20/00: Machine learning
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Various embodiments of the present disclosure provide machine learning configuration techniques for seamlessly leveraging compute functionalities from across a plurality of disparate third-party computing resources. The configuration techniques include receiving a first-party workspace request that identifies a third-party computing resource and in response to the first-party workspace request: generating a compute agnostic project workspace hosted by a first-party computing resource, initiating the generation of a third-party workspace hosted by the third-party computing resource, and initiating the configuration of a first-party routine set within the third-party workspace. The first-party routine set includes a plurality of webhooks that facilitate communication between the first-party computing resource and the third-party computing resource, thereby enabling a first-party to leverage multiple different, traditionally incompatible, computing functionalities from one centralized location.

## Claims

### Claim 1

A computer-implemented method performed by a first-party computing resource configured to facilitate a first stage of a machine learning project, the computer-implemented method comprising:
receiving, by one or more processors of the first-party computing resource, a first-party workspace request that is indicative of a third-party computing resource for facilitating a second stage of the machine learning project; and
in response to the first-party workspace request:
generating, by the one or more processors, a compute agnostic project workspace hosted by the first-party computing resource that is configured to access a third-party workspace hosted by the third-party computing resource,
initiating, by the one or more processors, the generation of the third-party workspace, and
initiating, by the one or more processors, the configuration of a first-party routine set within the third-party workspace, wherein the first-party routine set comprises a plurality of callbacks for facilitating communication between the first-party computing resource and the third-party computing resource.

### Claim 2 (depends on 1)

The computer-implemented method of claim 1 further comprising:
receiving, through the first-party routine set, a publication request from the third-party computing resource; and
in response to the publication request:
generating evaluation data for a portion of the machine learning project within the third-party workspace by initiating the performance of one or more project quality routines from the first-party routine set within the third-party workspace, and
modifying the compute agnostic project workspace based on a comparison between the evaluation data and one or more publication criteria.

### Claim 3 (depends on 2)

The computer-implemented method of claim 2, wherein modifying the compute agnostic project workspace comprises:
in response to determining that the machine learning project satisfies the one or more publication criteria, generating a canonical representation of the machine learning project that comprises one or more model attributes for the machine learning project.

### Claim 4 (depends on 3)

The computer-implemented method of claim 3, wherein the one or more model attributes comprise one or more model quality metrics for the machine learning project.

### Claim 5 (depends on 4)

The computer-implemented method of claim 4, wherein the one or more publication criteria comprises one or more model quality thresholds indicative of an acceptable publication threshold for each of the one or more model quality metrics.

### Claim 6 (depends on 1)

The computer-implemented method of claim 1, wherein the first-party workspace request is indicative of a plurality of third-party computing resources, and wherein the computer-implemented method further comprises:
initiating the generation of a respective third-party workspace for each of the plurality of third-party computing resources; and
initiating the configuration of the first-party routine set within the respective third-party workspace for each of the plurality of third-party computing resources to facilitate communication between the first-party computing resource and each of the plurality of third-party computing resources.

### Claim 7 (depends on 1)

The computer-implemented method of claim 1, wherein receiving the first-party workspace request comprises:
receiving a first selection input from a first selection interface hosted by the first-party computing resource, wherein the first selection input identifies a first third-party computing resource for configuring a training dataset for the machine learning project; and
receiving a second selection input from a second selection interface hosted by the first-party computing resource, wherein the second selection input identifies a second third-party computing resource for configuring a model for the machine learning project.

### Claim 8 (depends on 7)

The computer-implemented method of claim 7 further comprising:
providing the first selection interface for display, wherein the first selection interface comprises one or more interactive data selection widgets indicative of a first plurality of third-party computing resources for data configuration; and
providing the second selection interface for display, wherein the second selection interface comprises one or more interactive compute selection widgets indicative of a second plurality of third-party computing resources for model configuration.

### Claim 9 (depends on 7)

The computer-implemented method of claim 7, wherein the first third-party computing resource and the second third-party computing resource are different third-party computing resources.

### Claim 10 (depends on 1)

The computer-implemented method of claim 1, wherein the compute agnostic project workspace comprises a plurality of sub-workspaces for the machine learning project, wherein each sub-workspace is associated with one of a plurality of stages of development for the machine learning project.

### Claim 11 (depends on 10)

The computer-implemented method of claim 10, wherein the plurality of stages of development comprises a data preparation stage, a model experiment stage, a model review stage, and a model deployment stage.

### Claim 12 (depends on 10)

The computer-implemented method of claim 10, wherein a first sub-workspace and the third-party workspace are associated with the first stage of the plurality of stages of development, and wherein the computer-implemented method further comprises:
generating a first project interface corresponding to the first stage, wherein the first project interface comprises one or more interactive third-party links to the third-party workspace.

### Claim 13 (depends on 12)

The computer-implemented method of claim 12 further comprising:
receiving user input indicative of a selection of at least one of the one or more interactive third-party links; and
in response to the user input, initiating, through the first-party routine set, the performance of a computing action at the third-party workspace.

### Claim 14 (depends on 13)

The computer-implemented method of claim 13, wherein the computing action comprises a publication request for publishing at least a portion of the machine learning project.

### Claim 15 (depends on 13)

The computer-implemented method of claim 13, wherein the computing action comprises an access request for accessing the third-party workspace.

### Claim 16 (depends on 1)

The computer-implemented method of claim 1, wherein the first-party workspace request is associated with one or more user subscriptions, and wherein the third-party workspace is generated using the one or more user subscriptions.

### Claim 17 (depends on 16)

The computer-implemented method of claim 16, wherein the one or more user subscriptions are indicative of one or more resource permissions for the third-party computing resource, wherein the one or more resource permissions are indicative of at least one of an amount of allocated space for a user or an amount of allocated compute power for the user.

### Claim 18

A computing system of a first-party computing resource configured to facilitate a first stage of a machine learning project, the computing system comprising memory and one or more processors communicatively coupled to the memory, the one or more processors configured to:
receive a first-party workspace request that is indicative of a third-party computing resource for facilitating a second stage of the machine learning project; and
in response to the first-party workspace request:
generate a compute agnostic project workspace hosted by the first-party computing resource that is configured to access a third-party workspace hosted by the third-party computing resource,
initiate the generation of the third-party workspace, and
initiate the configuration of a first-party routine set within the third-party workspace, wherein the first-party routine set comprises a plurality of callbacks for facilitating communication between the first-party computing resource and the third-party computing resource.

### Claim 19 (depends on 18)

The computing system of claim 18, wherein the first-party workspace request is indicative of a plurality of third-party computing resources, and wherein the one or more processors are further configured to:
initiate the generation of a respective third-party workspace for each of the plurality of third-party computing resources; and
initiate the configuration of the first-party routine set within the respective third-party workspace for each of the plurality of third-party computing resources to facilitate communication between the first-party computing resource and each of the plurality of third-party computing resources.

### Claim 20

One or more non-transitory computer-readable storage media including instructions that, when executed by one or more processors of a first-party computing resource configured to facilitate a first stage of a machine learning project, cause the one or more processors to:
receive a first-party workspace request that is indicative of a third-party computing resource for facilitating a second stage of the machine learning project; and
in response to the first-party workspace request:
generate a compute agnostic project workspace hosted by the first-party computing resource that is configured to access a third-party workspace hosted by the third-party computing resource,
initiate the generation of the third-party workspace, and
initiate the configuration of a first-party routine set within the third-party workspace, wherein the first-party routine set comprises a plurality of callbacks for facilitating communication between the first-party computing resource and the third-party computing resource.
