# US 2026/0099305 A1 · Systems and methods for universal machine learning pipeline execution

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

- Number: US 2026/0099305 A1
- Document ID: US20260099305A1
- Status: Published application
- Kind: A1
- Application number: 19/414,751
- Filed: 2025-12-10
- Published: 2026-04-09
- Earliest claimed: 2023-09-11
- Assignees: Wells Fargo Bank, N.A.
- Applicants: Wells Fargo Bank, N.A.
- Inventors: Rameshchandra Bhaskar Ketharaju; Anjeet Kumar; Shuvam Sengupta
- Web page: https://patentgrep.com/patent/US20260099305A1

## CPC

- G06F 8/35: model driven
- G06F 8/31: {Programming languages or programming paradigms}
- G06F 11/302: {where the computing system component is a software system}
- G06F 11/3093: {Configuration details thereof, e.g. installation, enabling, spatial arrangement of the probes}
- G06F 11/3409: {for performance assessment}
- G06F 11/3466: {Performance evaluation by tracing or monitoring}
- G06F 11/3476: {Data logging (G06F11/14, G06F11/2205 take precedence)}
- G06N 20/00: Machine learning
- G06F 2201/865: Monitoring of software
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

Systems, apparatuses, methods, and computer program products are disclosed for automated model development. An example method includes parsing, by configuration circuitry, a configuration file, and generating, by an execution engine and based on the parsed configuration file, model code for training and testing a machine learning model. The example method further includes generating, by the execution engine, a machine learning pipeline, wherein the machine learning pipeline comprises the model code and a data processing engine, and instantiating, by a monitoring driver, a monitoring engine to monitor the machine learning pipeline. The example method further includes causing execution, by the execution engine, of the machine learning pipeline, and during execution of the machine learning pipeline, generating model performance data by the monitoring engine. The example method further includes receiving, by from the execution engine, model output data, wherein the model output data comprises a trained model and model performance data.

## Claims

### Claim 1

A method for machine learning model monitoring, the method comprising:
parsing, by configuration circuitry, a configuration file indicating a machine learning model belonging to a machine learning pipeline;
instantiating, by a monitoring driver, a monitoring engine to monitor the machine learning pipeline;
detecting an execution, by an execution engine, of the machine learning pipeline;
during the execution of the machine learning pipeline:
detecting, by the monitoring engine, a modification of a parameter of the machine learning model,
determining, by the monitoring engine and based on the modification of the parameter of the machine learning model, a violation of a model standardization condition, and
generating, by the monitoring engine, model performance data in real time simultaneously with the execution of the machine learning pipeline, wherein the model performance data comprises an indication of the violation of the model standardization condition; and
receiving, by communications hardware and from the execution engine, model output data.

### Claim 2 (depends on 1)

The method of claim 1, further comprising:
accessing, by a data processing engine, a data store;
retrieving, by the data processing engine, a data feature from the data store; and
providing, by the data processing engine, the data feature to the machine learning pipeline.

### Claim 3 (depends on 2)

The method of claim 2, wherein the data store comprises a low-latency data source, and wherein providing the data feature to the machine learning pipeline comprises:
staging, by the data processing engine, the data feature to a low-latency memory.

### Claim 4 (depends on 2)

The method of claim 2, wherein the data store comprises a cloud data source, and wherein providing the data feature to the machine learning pipeline comprises:
staging, by the data processing engine, the data feature to a local memory from the cloud data source.

### Claim 5 (depends on 1)

The method of claim 1, wherein the configuration file comprises an indication of a programming language, and wherein the machine learning pipeline comprises model code comprising computer instructions written in the programming language.

### Claim 6 (depends on 1)

The method of claim 1, wherein the configuration file comprises an indication of the model standardization condition, and wherein the method further comprises:
recording, by the monitoring engine and to a monitoring log, violation data describing the violation of the model standardization condition.

### Claim 7 (depends on 1)

The method of claim 1, wherein:
the configuration file comprises an indication of a model environment;
the indication of the model environment is drawn from a list of model environments;
the list of model environments comprises training, testing, and production; and
causing the execution of the machine learning pipeline uses the model environment.

### Claim 8 (depends on 1)

The method of claim 1, wherein receiving the model output data comprises:
storing, by the communications hardware, the model output data on a cloud data store, wherein the configuration file comprises an indication of storing the model output data to the cloud data store.

### Claim 9

An apparatus for automated model development, the apparatus comprising:
configuration circuitry configured to:
parse a configuration file indicating a machine learning model belonging to a machine learning pipeline;
a monitoring driver configured to:
instantiate a monitoring engine to monitor the machine learning pipeline;
an execution engine configured to:
detect an execution of the machine learning pipeline,
wherein the monitoring engine is further configured to, during the execution of the machine learning pipeline:
detect a modification of a parameter of the machine learning model,
determine, based on the modification of the parameter of the machine learning model, a violation of a model standardization condition, and
generate model performance data in real time simultaneously with the execution of the machine learning pipeline, wherein the model performance data comprises an indication of the violation of the model standardization condition; and
communications hardware configured to:
receive, from the execution engine, model output data.

### Claim 10 (depends on 9)

The apparatus of claim 9, wherein the data processing engine is configured to:
access a data store;
retrieve a data feature from the data store; and
provide the data feature to the machine learning pipeline.

### Claim 11 (depends on 10)

The apparatus of claim 10, wherein the data store comprises a low-latency data source, and wherein the data processing engine is configured such that providing the data feature to the machine learning pipeline comprises:
staging the data feature to a low-latency memory.

### Claim 12 (depends on 10)

The apparatus of claim 10, wherein the data store comprises a cloud data source, and wherein the data processing engine is configured such that providing the data feature to the machine learning pipeline comprises:
staging the data feature to a local memory from the cloud data source.

### Claim 13 (depends on 9)

The apparatus of claim 9, wherein the configuration file comprises an indication of a programming language, and wherein the machine learning pipeline comprises model code comprising computer instructions written in the programming language.

### Claim 14 (depends on 9)

The apparatus of claim 9, wherein the configuration file comprises an indication of the model standardization condition, and wherein the monitoring engine is further configured to:
record, to a monitoring log, violation data describing the violation of the model standardization condition.

### Claim 15 (depends on 9)

The apparatus of claim 9, wherein:
the configuration file comprises an indication of a model environment,
the indication of the model environment is drawn from a list of model environments,
the list of model environments comprises training, testing, and production, and
causing the execution of the machine learning pipeline uses the model environment.

### Claim 16 (depends on 9)

The apparatus of claim 9, wherein the communications hardware is further configured such that receiving the model output data comprises storing the model output data on a cloud data store, and wherein the configuration file comprises an indication of storing the model output data to the cloud data store.

### Claim 17

An apparatus for automated model development, the apparatus comprising:
means for parsing a configuration file indicating a machine learning model belonging to a machine learning pipeline;
means for instantiating a monitoring engine to monitor the machine learning pipeline;
means for detecting an execution of the machine learning pipeline;
means for, during the execution of the machine learning pipeline:
detecting a modification of a parameter of the machine learning model,
determining, based on the modification of the parameter of the machine learning model, a violation of a model standardization condition, and
generating model performance data in real time simultaneously with the execution of the machine learning pipeline, wherein the model performance data comprises an indication of the violation of the model standardization condition; and
means for receiving model output data.

### Claim 18 (depends on 17)

The apparatus of claim 17, further comprising:
means for accessing a data store;
means for retrieving a data feature from the data store; and
means for providing the data feature to the machine learning pipeline.

### Claim 19 (depends on 18)

The apparatus of claim 18, wherein the data store comprises a low-latency data source, and wherein means for providing the data feature to the machine learning pipeline comprises:
staging, by the data processing engine, the data feature to a low-latency memory.

### Claim 20 (depends on 18)

The apparatus of claim 18, wherein the data store comprises a cloud data source, and wherein means for providing the data feature to the machine learning pipeline comprises:
staging, by the data processing engine, the data feature to a local memory from the cloud data source.
