Jobs / Data and analytics
Can agents do data and analytics work?
Query, clean, analyze, and model data; statistics and operations research.
- US jobs
- 499,540 (0.3% of employment)
- Typical wage
- $115K a year (employment-weighted mean of occupation medians)
- Occupations
- 13 in O*NET 31.0
- Evidence
- 2 direct benchmarks, 4 also relevant, 4 context only
The answer today
Some tasksNo change since 10 Oct 2025.
Rule: Agents complete 20% or more of the tasks (or projects) on a direct benchmark.
The evidence that sets this step
- Spider 2.0 by HKU and Salesforce AI Research: 96.7% of tasks completed · Genloop's Sentinel Agent v2 Pro, 1 Mar 2026
Next step: Most tasks needs 60% or more of the tasks (or projects) on two direct benchmarks from independent institutions.
How much of the work is measured
8% of the work in this family is measured by direct benchmarks. Little of the work is measured (less than 20%). On 10 Oct 2025: 8%.
Work = the O*NET 31.0 work activities of the family's occupations, weighted by how important each task is and by US employment. A benchmark measures an activity when its tasks make the agent do it and its grading checks it. Method.
| Work activity (O*NET) | Share of the work | Direct benchmark |
|---|---|---|
| Develop operational or technical procedures or standards | 7.3% | Not measured |
| Analyze scientific or applied data using mathematical principles | 6.4% | Not measured |
| Present research or technical information | 6.3% | Not measured |
| Analyze market or industry conditions | 5.8% | Not measured |
| Analyze data to improve operations | 5.1% | Not measured |
| Maintain current knowledge in area of expertise | 4.2% | Not measured |
| Develop scientific or mathematical theories or models | 4.2% | Not measured |
| Process digital or online data | 4.1% | Measured |
| Design computer or information systems or applications | 3.9% | Not measured |
| Design databases | 3.6% | Not measured |
Our reading
Agents complete some data tasks. On Spider 2.0-Snow, the top entry answers 96.7% of 547 business questions with SQL against a large warehouse. The entry does not name its model, and scores are self-reported and then validated by the maintainers. When the request is ambiguous and the agent must ask a simulated user, success falls: 22.67% on BIRD-Interact, from a self-submitted entry that the maintainers have not verified. No direct benchmark measures analyzing data with mathematical methods or presenting research or technical information. A second benchmark from an independent institution at 60% or more would move the family to most tasks.
Evidence: Spider 2.0, BIRD-Interact by Google Cloud, HKU, and Salesforce AI Research. Written 24 Sep 2026.
Benchmarks for this work
Direct evidence
Work benchmarks mapped primary to this family. They set the step when they report how many tasks agents complete.
- BIRD-Interactby HKU and Google Cloud22.7%MERIT + GPT-5.4 (GPT-4o user simulator) · 22 May 2026Emerging
- 96.7%Genloop's Sentinel Agent v2 Pro · 1 Mar 2026Saturated
Also relevant
These benchmarks cover part of this family's work, or have no result yet. They do not set its step.
- 54.5%arito · 13 Sep 2026Emerging
Context only
These benchmarks span many occupations, or they test how agents behave. They give context but do not set the step.
Functions: Data engineering and SQLAsking for clarificationWeb researchMedical image readingClinical records workMCP tool useEnterprise workflowsML engineeringFreelance projectsProfessional deliverablesSpreadsheet workFinancial analysis
What we do not know
- Benchmarks measure tasks, not whole jobs. A job is a bundle of tasks, judgment, and relationships.
- Capability is not adoption. A task that agents can do in a benchmark may still be done by people at work.
- Most results come from single runs. Reliability over many runs is often unknown.
Occupations in this family
Sorted by US employment (BLS OEWS May 2025). O*NET sub-occupations share the employment figure of their SOC group. Work measured = the share of the occupation's O*NET work that direct benchmarks test.
| Occupation | SOC | US jobs | Work measured |
|---|---|---|---|
| Geographic Information Systems Technologists and Technicians | 15-1299.02 | (435K group) | 0% |
| Business Intelligence Analysts | 15-2051.01 | (262K group) | 13% |
| Clinical Data Managers | 15-2051.02 | (262K group) | 13% |
| Data Scientists | 15-2051.00 | 262K | 13% |
| Operations Research Analysts | 15-2031.00 | 109K | 0% |
| Data Warehousing Specialists | 15-1243.01 | (67K group) | 7% |
| Database Architects | 15-1243.00 | 67K | 7% |
| Biostatisticians | 15-2041.01 | (29K group) | 0% |
| Statisticians | 15-2041.00 | 29K | 0% |
| Actuaries | 15-2011.00 | 27K | 0% |
| Bioinformatics Technicians | 15-2099.01 | (4K group) | 0% |
| Mathematical Science Occupations, All Other | 15-2099.00 | 4K | 0% |
| Mathematicians | 15-2021.00 | 2K | 0% |
How to cite
Credit the original work first: the benchmarks listed on this page, and their institutions.
Then, if you used this page:
Can Agents Work. "Can agents do data and analytics work?." https://canagentswork.com/jobs/data-analytics/ (accessed 2026-10-11). CC BY 4.0.@misc{caw-jobs-data-analytics,
title = {{Can agents do data and analytics work?}},
author = {{Can Agents Work}},
year = {2026},
howpublished = {\url{https://canagentswork.com/jobs/data-analytics/}},
note = {Accessed 2026-10-11. CC BY 4.0}
}