# US 2025/0384352 A1 · Computing systems and methods for a unified machine learning pipeline with a logging adapter > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2025/0384352 A1 - Document ID: US20250384352A1 - Status: Published application - Kind: A1 - Application number: 18/745,540 - Filed: 2024-06-17 - Published: 2025-12-18 - Earliest claimed: 2024-06-17 - Applicants: The Toronto-Dominion Bank - Inventors: Peter STARSZYK; Anuar YERALIYEV; Satya Krishna GORTI; Keyvan GOLESTAN IRANI; Alexander CLARENCE; Kin Kwan LEUNG; Devinder KUMAR; Raunaq SURI; Guangwei YU; Maksims VOLKOVS - Web page: https://patentgrep.com/patent/US20250384352A1 ## CPC - G06N 20/20: Ensemble learning - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Systems and methods are provided for a machine learning (ML) pipeline with a unified framework. A machine learning pipeline trains a machine learning model in the machine learning pipeline in a development environment to generate training artifacts. The machine learning pipeline further executes the machine learning model in a production environment to generate production artifacts. When a training data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the training artifacts to the training data logger for storage in the development environment. When the production data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the production artifacts to the production data logger for storage in the production environment. ## Claims ### Claim 1 A cloud computing system for machine learning, the cloud computing system comprising: a machine learning pipeline configured to train a machine learning model in the machine learning pipeline in a development environment and to generate training artifacts, and further configured to execute the machine learning model in a production environment to generate production artifacts; a training data logger, wherein, when the training data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the training artifacts to the training data logger for storage in the development environment; and a production data logger, wherein, when the production data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the production artifacts to the production data logger for storage in the production environment. ### Claim 2 (depends on 1) The cloud computing system of claim 1, wherein the machine learning pipeline is configured to synchronize logged data from the development environment and logged data from the production environment; and wherein the logged data from the development environment comprises the training artifacts, and the logged data from the production environment comprises the production artifacts. ### Claim 3 (depends on 1) The cloud computing system of claim 1, wherein the training data logger transmits back the training artifacts to the machine learning pipeline for further training of the machine learning model in the development environment. ### Claim 4 (depends on 1) The cloud computing system of claim 1, wherein the training data logger stores the training artifacts to a database in the development environment. ### Claim 5 (depends on 4) The cloud computing system of claim 4, wherein the production data logger comprises memory, and the production data logger is configured to asynchronously store the production artifacts in the memory. ### Claim 6 (depends on 1) The cloud computing system of claim 1, wherein the training artifacts comprises intermediate data generated from training the machine learning model, or feature generation steps, or a trained model object, or a combination thereof. ### Claim 7 (depends on 6) The cloud computing system of claim 6, wherein the training artifacts comprises the trained model object, and the training data logger transmits back the trained model object to the machine learning pipeline for executing an inference process using the machine learning model. ### Claim 8 (depends on 1) The cloud computing system of claim 1, wherein the machine learning pipeline is further configured to execute the machine learning model in a batch inferencing environment to generate batch inference artifacts; and the cloud computing system further comprises a testing data logger and, when a test data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically determines that the batch inference artifacts are transmitted to the test data logger for storage in the batch inferencing environment. ### Claim 9 (depends on 1) The cloud computing system of claim 1, wherein the production environment is a real-time inferencing environment, and the production artifacts are real-time inferencing artifacts. ### Claim 10 (depends on 9) The cloud computing system of claim 9, further comprising an artifact consumer in the real-time inferencing environment; and wherein, in the real-time inferencing environment, the machine learning pipeline receives a real-time request, the machine learning pipeline processes the real-time request and generates the real-time inferencing artifacts, and the artifact consumer processes the real-time inferencing artifacts to generate a response to the real-time request. ### Claim 11 A method for machine learning, the method executed in a computing environment comprising one or more processors, a communication interface, and memory, and the method comprising: a machine learning pipeline training a machine learning model in the machine learning pipeline in a development environment to generate training artifacts, and further executing the machine learning model in a production environment to generate production artifacts; wherein, when a training data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the training artifacts to the training data logger for storage in the development environment; and, wherein, when a production data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the production artifacts to the production data logger for storage in the production environment. ### Claim 12 (depends on 11) The method of claim 11, further comprising the machine learning pipeline synchronizing logged data from the development environment and logged data from the production environment; and wherein the logged data from the development environment comprises the training artifacts, and the logged data from the production environment comprises the production artifacts. ### Claim 13 (depends on 11) The method of claim 11, further comprising the training data logger transmitting back the training artifacts to the machine learning pipeline for further training of the machine learning model in the development environment. ### Claim 14 (depends on 11) The method of claim 11, further comprising the training data logger storing the training artifacts to a database in the development environment. ### Claim 15 (depends on 14) The method of claim 14, wherein the production data logger comprises memory, and the method further comprising the production data logger asynchronously storing the production artifacts in the memory. ### Claim 16 (depends on 11) The method of claim 11, wherein the training artifacts comprises intermediate data generated from training the machine learning model, or feature generation steps, or a trained model object, or a combination thereof. ### Claim 17 (depends on 16) The method of claim 16, wherein the training artifacts comprises the trained model object, and the method further comprising the training data logger transmits back the trained model object to the machine learning pipeline for executing an inference process using the machine learning model. ### Claim 18 (depends on 11) The method of claim 11, further comprising: the machine learning pipeline executing the machine learning model in a batch inferencing environment to generate batch inference artifacts; and, when a testing data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically determines that the batch inference artifacts are transmitted to the testing data logger for storage in the batch inferencing environment. ### Claim 19 (depends on 11) The method of claim 11, wherein the production environment is a real-time inferencing environment, and the production artifacts are real-time inferencing artifacts, and the method further comprising: in the real-time inferencing environment, the machine learning pipeline receives a real-time request; the machine learning pipeline processing the real-time request and generating the real-time inferencing artifacts; and an artifact consumer processing the real-time inferencing artifacts to generate a response to the real-time request. ### Claim 20 A non-transitory computer readable medium storing computer executable instructions which, when executed by at least one computer processor, cause the at least one computer processor to carry out a method for machine learning, the method comprising: a machine learning pipeline training a machine learning model in the machine learning pipeline in a development environment to generate training artifacts, and further executing the machine learning model in a production environment to generate production artifacts; wherein, when a training data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the training artifacts to the training data logger for storage in the development environment; and, wherein, when a production data logger and the machine learning pipeline are in communication with each other, the machine learning pipeline automatically transmits the production artifacts to the production data logger for storage in the production environment.