SOC 15-2021 · Job decomposition

Mathematicians

What this job actually does, classified into Automate, Augment, and Human-only. Decomposed from 29 canonical roles mapped to this SOC across Nuvepro's real-world JD corpus.

Tasks drawn from the broader SOC minor group 15-20% because the exact SOC has thin coverage in our corpus.

2,030

U.S. employment

BLS OEWS

$126,710

Median annual wage

BLS OEWS

42%

AEI exposure

Anthropic Economic Index

774

Distinct tasks

From 29 roles

Task mix (by distinct task)

Automate · 26 (3%)

Augment · 612 (79%)

Human-only · 136 (18%)

Automate

12 of 26

AI does it end-to-end

Clean data for analysis.
18 jobs · 2 roles · 100% confidence
Monitor models in production.
14 jobs · 1 role · 71% confidence
Retrain models in production.
14 jobs · 1 role · 86% confidence
Query APIs to collect data.
12 jobs · 1 role · 100% confidence
Clean data to prepare it for analysis.
10 jobs · 1 role · 80% confidence
Clean and condition data.
9 jobs · 1 role · 67% confidence
Cleanse data to prepare it for analysis.
9 jobs · 1 role · 100% confidence
Validate incoming data against established standards and criteria.
6 jobs · 1 role · 83% confidence
Automate reporting processes for regular and ad-hoc stakeholder needs.
5 jobs · 1 role · 60% confidence
Cleanse data for analysis.
5 jobs · 1 role · 100% confidence
Execute tools to report on data quality.
5 jobs · 1 role · 100% confidence
Retrieve data from a variety of sources and structures
5 jobs · 1 role · 100% confidence

Augment

12 of 612

Human + AI together

Analyze data-rich environments to identify relevant information.
65 jobs · 2 roles · 100% confidence
Develop predictive data models.
44 jobs · 2 roles · 100% confidence
Provide a deep understanding of data and its meaning.
32 jobs · 1 role · 97% confidence
Use appropriate tools and frameworks to transform disparate data points into objective answers.
26 jobs · 2 roles · 100% confidence
Create schemas for data storage.
18 jobs · 1 role · 100% confidence
Create ETL pipelines.
17 jobs · 1 role · 100% confidence
Monitor and evaluate the performance of deployed models.
16 jobs · 1 role · 94% confidence
Ensure data quality and integrity in all processes.
15 jobs · 1 role · 100% confidence
Stay updated with the latest trends in data science.
15 jobs · 1 role · 100% confidence
Develop algorithms to solve client problems.
14 jobs · 2 roles · 100% confidence
Develop and optimize machine learning models for various applications.
14 jobs · 1 role · 100% confidence
Lead the development and implementation of advanced data science models.
14 jobs · 1 role · 86% confidence

Human-only

12 of 136

Judgment, taste, accountability

Work closely with clients to understand their questions and needs.
58 jobs · 2 roles · 100% confidence
Guide teammates in data science work.
32 jobs · 2 roles · 100% confidence
Lead the development of algorithms and systems.
30 jobs · 2 roles · 74% confidence
Work with clients to understand their questions and data needs.
20 jobs · 1 role · 100% confidence
Advise clients using data-driven findings to support informed decisions.
17 jobs · 1 role · 71% confidence
Drive best practices in data science.
16 jobs · 1 role · 94% confidence
Mentor and guide junior data scientists.
16 jobs · 1 role · 100% confidence
Collaborate with stakeholders to understand requirements.
15 jobs · 1 role · 93% confidence
Collaborate with cross-functional teams to integrate ML models into products and services.
13 jobs · 1 role · 100% confidence
Govern models from a risk perspective.
13 jobs · 1 role · 100% confidence
Cross-train as workstation controller when feasible.
12 jobs · 1 role · 100% confidence
Lead the development of solutions to complex programs.
11 jobs · 2 roles · 100% confidence

Audit this role for your org

See which of these tasks your team is doing today, and how to ship the first AI-enabled version in 14 days.

Audit your team →Browse all tasks →