Data Analyst

What you'll do

The project evaluates AI assistants doing the kind of analytics work a data team does every week: investigating an anomaly, producing a KPI report, refreshing a dataset. You run these workflows through AI tools that are connected to a cloud data warehouse, and then you write and run your own SQL against the source data to see whether the assistant's numbers are right. Wrong joins, a filter that silently drops rows, a time window that is off by a day: those are the errors you are there to catch.

There is a real operations side as well. You keep the seeded Snowflake datasets in a known state, resetting them so each test scenario has a correct answer to compare against, and you manage the warehouse roles, grants and access controls that make the evaluation environment secure and repeatable. You set up and document how the assistants authenticate to multiple analytics and SaaS platforms, and when a product does something undocumented you write it up. Calibration sessions with other evaluators keep the grading consistent.

Who fits

• At least 3 years as a data analyst or analytics engineer
• Advanced SQL and hands-on Snowflake: warehouses, access control, query history
• A habit of reconciling reported metrics against raw data and spotting aggregation or logic errors
• Experience using connectors in Claude and ChatGPT
• Grounding in business and finance analytics, including KPI definitions for leadership or external reporting
• Database administration experience: roles, grants, authentication; OAuth or security integrations are a plus
• Working knowledge of Slack, Google Workspace or Microsoft 365 for sharing results
• A critical eye on AI-generated SQL from your own use of AI assistants
• Rubric-based evaluation, QA or data labelling experience, and clear communication

These are listed as preferred qualifications, but a Snowflake admin and SQL audit background is plainly what the work needs.

What it pays

$50–60/hour, micro1's published figure. It is a narrow band, so the number is close to what you will be offered. The ad publishes no compensation structure, weekly hours or project length, so whether this is paid for time or per task is not stated.

Worth knowing

Good:

• Real analytics engineering work
• A narrow, predictable band
• Snowflake administration and connector setup are skills that transfer straight back to industry
• No AI experience required beyond using assistants critically

Less good:

• Only 4 openings, and the post has been up about a month
• $50–60 is modest for someone with Snowflake admin and security-integration experience
• No weekly hours, project length or pay structure stated
• The client is described only as confidential
• Contractor engagement: no benefits, your own tax to manage

About this listing

Posted by micro1 on its own jobs portal and read on 24 September 2026, about a month after it went up. The $50–60/hour band, the 4 openings and the qualifications are its published text. Location is Remote with no country list, and payment schedule and method are not published on the listing. See micro1.