Stakeholders and Terminologies Used in MLOPS

Overview

MLOps (Machine Learning Operations) is a methodology that involves various stakeholders and terminologies. Here is an overview of the key stakeholders and terminologies used in MLOps.

Stakeholders in MLOPS

The stakeholders of AI/MLOps include:

• Subject Matter Expert(SME): An SME will ask business questions and ensure model performance and meet our business needs or goals.
• Data Analysts: This person handles and does the exploratory data analysis. He will help with developing features for the ML model.
• Data Scientists: Data scientists are responsible for developing and deploying AI/ML models.
• Software Engineers: Software engineers are responsible for developing and maintaining the underlying infrastructure for AI/MLOps.
• Business Executives: Business executives oversee the implementation and use of AI/MLOps.
• Product Managers: Product managers are responsible for driving the implementation and usage of AI/MLOps.
• IT Operations: IT operations manage and monitor the AI/MLOps environment.
• Security Professionals: Security professionals are responsible for ensuring the security of the AI/MLOps environment.

How would the terminologies help the students?

• It makes a difference in understanding the complexities of machine learning operations.
• They give a common dialect that can be utilized to communicate between distinctive groups and partners. They also provide an understanding of the various components of MLOps and how they interact.
• Finally, they provide the structure and guidelines for implementing MLOps in a way that is most efficient and effective for them.

The professor wanted the students to understand them because MLOps terminologies can help solve by streamlining the building, testing, deploying, and monitoring of machine learning models.

These terminologies enable teams to collaborate more effectively, automate processes for faster development, and increase visibility and control over the models in production.

MLOps terminologies also provide a standard approach to developing, deploying, and managing machine learning models, reducing manual effort and improving system reliability.

MLOps terminologies:

• Machine Learning (ML): Machine learning is a subfield of artificial intelligence (AI) concerned with developing systems that can learn from data and improve their performance over time.
• Artificial Intelligence (AI): Artificial intelligence (AI) is the ability of a computer program or a machine to think and learn. It is additionally a field of study which endeavors to form computers cleverly.
• Data Science: Data science is an interdisciplinary field that combines computer science, statistics, and other areas of study to extract meaningful insights from data.
• Deep Learning (DL): Deep learning is a subset of machine learning that utilizes artificial neural networks to process data and make predictions.
• Natural Language Processing (NLP): Natural language processing (NLP) is a subfield of artificial intelligence that focuses on enabling machines to understand and process natural language.
• Cloud Computing: Cloud computing is a type of computing that relies on sharing computing resources over a network, usually the Internet.
• DevOps: DevOps is a software development and operations methodology that emphasizes collaboration between software developers and system operators.
• CI/CD (Continuous Integration/Continuous Delivery): CI/CD is a software engineering approach in which developers regularly integrate code changes into a shared repository. Automated processes build, test, and deploy the application or system into production.
• Model Governance: Model Governance is managing, monitoring, and controlling the deployment, use, and maintenance of machine learning models in production.
• Model Validation: Model Validation ensures that a machine learning model is fit for its purpose and is functioning correctly.
• Model Monitoring: Model Monitoring is the process of continually tracking and analyzing the performance of a machine learning model in production.
• Data Version Control: Data Version Control is tracking and managing changes to data used to train and deploy machine learning models.
• Model Management: MLOps Model Management manages the entire lifecycle of machine learning models, from development and deployment to monitoring and maintenance. It includes mechanizing and streamlining the workflows to guarantee reliable and productive execution of models in generation.
• Flask: Flask is a lightweight Python web framework that enables the rapid development of web applications by providing valuable tools and features for building web applications. Flask is used for creating web applications using the Python programming language.
• Docker: Docker is a tool designed to make it easier to create, deploy, and run applications by using containers, which are lightweight, portable, and self-sufficient environments that can run on any host.
• Rest API: REST (Representational State Transfer) API is a set of rules and conventions for building web services that can be accessed over the Internet using standard HTTP methods.
• Fast API: FastAPI is a modern, fast (high-performance) web framework for building APIs with Python 3.6+ based on standard Python-type hints.
• GitHub: GitHub is a web-based platform for version control and collaboration that allows developers to store, share, and manage their code and projects.
• AWS ECS: Amazon Elastic Container Service (ECS) is a fully managed service that makes it easy to run, scale, and orchestrate Docker containers on AWS.
• IaaS: IaaS (Infrastructure as a Service) is a type of cloud computing that provides users with on-demand access to virtualized computing resources such as servers, storage, and networking.
• Streamlit: Streamlit is an open-source Python library for building and deploying custom machine learning and data science applications quickly and easily with an interactive user interface.

Conclusion

MLOps includes different stakeholders counting Subject Matter Specialists, Data Analysts, Data Scientists, Computer program Engineers, Business Executives, Item Supervisors, IT Operations, and Security Experts. MLOps wordings such as Machine Learning, Artificial Intelligence, Data Science, and DevOps give a system for understanding the complexities of machine learning operations.

Key Takeaways

1. MLOps includes numerous stakeholders such as subject matter specialists, data analysts, information researchers, program engineers, business executives, product managers, IT operations, and security experts.
2. Understanding MLOps phrasings can give a system for understanding the complexities of machine learning operations, a common dialect for communication between diverse groups and partners, and rules for actualizing MLOps in an proficient and compelling way.
3. Some key MLOps terminologies include machine learning, artificial intelligence, data science, deep learning, natural language processing, cloud computing, DevOps, CI/CD, model governance, model validation, model monitoring, data version control, model management, Flask, Docker, REST API, FastAPI, GitHub, AWS ECS, IaaS, and Streamlit.
4. MLOps terminologies can help streamline the building, testing, deploying, and monitoring of machine learning models, enable teams to collaborate more effectively, automate processes for faster development, and increase visibility and control over the models in production.

Quiz

1. Which stakeholder is responsible for developing and deploying AI/ML models?  
   1. SME  
   2. Data Analyst 
   3. Data Scientist
   4. Business Executive

Answer: c. Data Scientist

2. Which of the following is a subfield of artificial intelligence that focuses on enabling machines to understand and process natural language?  
   1. Cloud Computing  
   2. Deep Learning 
   3. Natural Language Processing 
   4. Model Governance

Answer: c. Natural Language Processing

3. What is DevOps? 
   1. A subset of machine learning that utilizes artificial neural networks to process data and make predictions.  
   2. A type of computing that relies on sharing computing resources over a network, usually the Internet.  
   3. A software development and operations methodology that emphasizes collaboration between software developers and system operators. 
   4. A set of rules and conventions for building web services that can be accessed over the Internet using standard HTTP methods.

Answer: c. A software development and operations methodology that emphasizes collaboration between software developers and system operators.

4. Which of the following is a lightweight Python web framework that enables the rapid development of web applications?  
   1. Flask 
   2. Docker  
   3. Rest API 
   4. Fast API

Answer: a. Flask