SQL Developer: Create Gold Queries for Sports Database

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help people start and grow careers teaching AI by connecting contributors to real projects, tools, and teams.

• Work remotely and build experience in the fast-growing human side of AI.
• OpenTrain projects provide access to data, tooling, and team support required to deliver high-quality labels and annotations.

About AI training work

AI training (data labeling/annotation) is how modern AI systems learn from examples prepared and reviewed by people. Tasks include creating, checking, and refining structured outputs — like SQL queries — so models can learn to turn human questions into correct database results.

This work is highly accessible, often flexible and remote, and gives contributors direct influence on how state-of-the-art systems behave.

• Typical advantages: remote work, flexible hours, and the chance to shape AI behavior.
• Many projects require only domain knowledge and attention to detail; specialist tasks reward deeper expertise.

The role

You will create gold-standard SQL queries for 100 user queries about sports, using a provided database schema and access to the actual DB for testing. Each SQL must return the correct results for its corresponding user question.

The project is contract, part-time work: expect 20+ hours/week and availability during East Coast hours for the project's scheduled window (Monday 24th through Wednesday 26th). You will use the team's labeling tool and work directly with existing teammates who will share contextual knowledge.

• Deliverables: 100 tested, well-documented SQL queries that answer the given user queries.
• Tooling: Access to the database and a labeling tool will be provided; collaboration and context are available from the existing team.

What you'll do day-to-day

Translate each natural-language user query into a precise, efficient SQL statement that returns the intended result from the provided schema. Test queries against the database and refine until they return correct results for edge cases.

Document assumptions, join logic, filters, and any schema nuances so reviewers understand your approach. Use the labeling tool to submit gold queries and collaborate with reviewers for clarifications.

• Write advanced SQL: joins, window functions, subqueries, aggregates, CTEs, and performance-aware constructs as needed.
• Test queries against the DB and iterate until they reliably produce correct results.
• Record notes and rationale for each query to support team review and future reuse.

Requirements

This role requires strong SQL expertise and hands-on experience producing advanced queries. You must be a native or highly fluent English speaker and have prior experience in data labeling or annotation tasks.

Availability for the specified short project window during East Coast hours is essential, and you must be able to commit 20+ hours per week for the duration.

• Expert-level SQL knowledge, including advanced queries and query optimization.
• Experience with data labeling or annotation workflows and attention to detail.
• Native or highly fluent English.
• Availability Monday 24th through Wednesday 26th during East Coast hours; 20+ hours/week.
• Subject matter: sports-related queries (domain familiarity helpful).

Schedule, pay, and logistics

This is a contract, part-time role billed hourly at $40 USD per hour. Candidates may work remotely from anywhere in the world, but must align with East Coast hours during the project window.

You will receive access to the database, the team's labeling tool, and direct support from the existing team for necessary context and tribal knowledge.

• Compensation: $40/hour, paid as a contractor.
• Employment types: CONTRACTOR, PART_TIME.
• Data type: TEXT; label type: TEXT_GENERATION; labeling software: provided (OTHER).

Who should apply and how it works

Apply if you are an SQL expert who enjoys translating human questions into exact database queries and has experience with data labeling workflows. Sports domain familiarity is a plus but not required.

Successful applicants will be given the dataset, DB access, and labeling tool, then work with the team to produce and submit the 100 gold queries within the agreed timeline.

• Ideal background: SQL developers, data engineers, or analysts with annotation experience.
• You must be prepared to test thoroughly and document each query for reviewer acceptance.