21211 Data scientists

What this occupation covers

Data scientists use advanced analytics technologies, including machine learning and predictive modelling, to support the identification of trends, scrape information from unstructured data sources and provide automated recommendations. They are employed by consulting firms, universities, banks and information technology departments in the public and private sectors.

Official main duties

The NOC profile lists 8 main duties for this code.

• Implement cutting-edge techniques and tools in machine learning, deep learning and artificial intelligence to make data analysis more efficient
• Perform large-scale experimentation to identify hidden relationships between variables in large datasets
• Create advanced machine learning algorithms such as regression, simulation, scenario analysis, modeling, clustering, decision trees and neural networks
• Prepare and extract data using programming language
• Implement new statistical, machine learning, or other mathematical methodologies to solve specific business problems
• Visualize data in a way that allows a business to quickly draw conclusions and make decisions
• Develop artificial intelligence models and algorithms and implement them to meet the needs of the organization
• Coordinate research and analysis activities using unstructured and structured data and use programming to clean and organize data

Do not copy these lines into your reference letter. The letter has to describe what you actually did, in your employer’s words. DUTYPROOF flags wording lifted from the NOC as something to review, and an officer reading a letter that mirrors the NOC reads it the same way.

What DUTYPROOF screens for

This is DUTYPROOF’s screening threshold, not IRCC’s decision. For this NOC, screening requires at least 2 of its 8 official duties to be evidenced in your own words, plus the required letter elements. IRCC asks that you performed the actions in the occupation’s lead statement and a substantial number of its main duties — so treat the duty coverage as the number that matters, not the badge.

Codes this one is confused with

The NOC profile names these itself. Choosing the wrong code of a look-alike pair is the most common way genuine work experience fails to establish eligibility.

• 20012 Computer and information systems managers
• 21311 Computer engineers (except software engineers and designers)
• 21223 Database analysts and data administrators
• 21222 Information systems specialists
• 21210 Mathematicians, statisticians and actuaries
• 21231 Software engineers and designers

Excluded from this code

• Computer and information systems managers (See 20012 Computer and information systems managers)
• Computer engineers (except software engineers and designers) (See 21311 Computer engineers (except software engineers and designers))
• Database analysts and data administrators (See 21223 Database analysts and data administrators)
• Information systems specialists (See 21222 Information systems specialists)
• Mathematicians, statisticians and actuaries (See 21210 Mathematicians, statisticians and actuaries)
• Software engineers and designers (See 21231 Software engineers and designers)

Job titles filed under this code

A title on this list does not by itself put your experience in this code — the duties in your letter decide that.

artificial intelligence (AI) analyst · artificial intelligence (AI) consultant · artificial intelligence (AI) designer · artificial intelligence analyst · artificial intelligence consultant · big data analyst · big data developer · big data engineer · data analytics consultant · data analytics specialist · data architect · data engineer · data miner · data mining analyst · data modeler · data scientist · lead data scientist · machine learning engineer · machine learning specialist · quantitative analyst

Employment requirements

• A bachelor's degree in statistics, mathematics, computer science, computer systems engineering or a related discipline or completion of a college program in computer science is usually required.
• A master's or doctoral degree in machine learning, data science, or a related quantitative field is usually required.
• Experience in programming is usually required.
• Experience in statistical modelling and machine learning is usually required.