AI Data Quality Analyst (Human-in-the-Loop)

We are looking for a hands-on AI Data Quality Analyst (Human-in-the-Loop) to support a strategic client in the commercial property insurance space. This role sits at the intersection of data quality, QA, and product thinking.

You will be the “human in the loop” for an AI-powered document processing pipeline: reviewing what the AI extracts from complex insurance submissions (e.g., Statements of Values (SOVs), loss runs, spreadsheets, PDFs), correcting errors, and ensuring that downstream tools receive clean, reliable data. On top of the day-to-day “grind work” of validation and correction, you’ll zoom out to identify recurring issues, spot patterns, and translate them into clear requirements and bug reports for the engineering team.

This is not a Product Manager role and not a “purely strategic” position. It is very hands-on, detail-oriented work that is critical to improving our AI systems and ultimately the client’s underwriting workflow.

• →Review AI-extracted data from insurance submissions (SOVs, loss runs, supporting documents) for accuracy, completeness, and consistency.
• →Compare extracted fields against source documents, identify discrepancies, and correct data directly in the appropriate systems or templates.
• →Act as a quality gate for the AI pipeline, ensuring output meets agreed business and underwriting expectations before it moves downstream.
• →Log issues, defects, and edge cases with clear reproduction steps, examples, and impact, using tools like Jira or similar.
• →Identify patterns and root causes behind extraction errors (e.g., recurring issues with specific formats, document types, or fields).
• →Translate observed patterns into well-structured requirements, user stories, and bug reports that engineering and data teams can act on.
• →Collaborate closely with architects, data engineers, and other analysts to refine extraction rules, templates, and workflows.
• →Use LLMs and AI assistants as tools (e.g., for summarization, cross-checking, hypothesis generation), while exercising sound judgment about what to trust and what to verify.
• →Help continuously improve documentation, checklists, and guidelines for reviewing submissions and extractions.
• →Over time, contribute to defining metrics and dashboards for data quality and model performance (e.g., accuracy by field, error rates by document type).
• →Work primarily in Eastern European time zones with sufficient overlap to collaborate with US-based stakeholders.

• 3+ years of experience in a data-intensive role such as data analyst, business analyst, QA analyst, operations analyst, or similar.
• Strong attention to detail and proven experience doing systematic, repetitive data review without loss of quality.
• Excellent analytical skills: ability to trace issues from symptoms (wrong numbers, missing fields) back to likely root causes (document patterns, parsing logic, business rules).
• Demonstrated ability to write clear, structured tickets/requirements for engineering teams (e.g., bug reports, user stories, acceptance criteria).
• Advanced Excel skills (pivot tables, lookups, filters, data cleansing techniques).
• Comfortable working with complex business documents and datasets (financial, insurance, or similar structured data).
• Strong written English for documenting findings, writing tickets, and communicating with distributed teams.
• Comfortable with “grind work”: reviewing many documents/records per day, while maintaining consistency and care.
• Experience with issue-tracking or project management tools (e.g., Jira, Azure DevOps, Trello, or similar).
• Ability to work independently, manage your own queue, and escalate appropriately when patterns or blockers emerge.

• Experience in commercial insurance, property insurance, or financial services operations.
• Familiarity with Statements of Values (SOVs), loss runs, or similar risk/coverage documentation.
• Exposure to LLMs and AI systems (e.g., document intelligence, RAG, chat agents) — as a user, tester, or collaborator.
• Basic familiarity with SQL or other query tools for validating data.
• Prior experience in a Human-in-the-Loop (HIL) or data quality role supporting machine learning models.
• Experience defining or working with data quality metrics (accuracy, completeness, precision/recall, etc.).

• Hands-on ownership: You’re happy to roll up your sleeves, dig into the data, and do what it takes to ensure quality.
• Curiosity & pattern recognition: You naturally look for trends behind individual issues and want to understand “why” things fail.
• Structured thinking: You can turn messy observations into clear, actionable tickets and requirements.
• Communication & collaboration: You’re comfortable working with engineers, architects, and business stakeholders across time zones.
• Adaptability: You’re eager to learn new tools, domains, and workflows in a rapidly evolving AI environment.

Nimble Gravity is a team of outdoor enthusiasts, adrenaline seekers, and experienced growth hackers. We love solving hard problems and believe the right data can transform and propel growth for any organization.

We work at the cutting edge of data, analytics, and AI, helping clients build and scale solutions that deliver meaningful business impact.

Nimble Gravity is an Equal Opportunity Employer and considers applicants for employment without regard to race, color, religion, sex, orientation, national origin, age, disability, genetics, or any other basis forbidden under applicable law. Nimble Gravity considers all qualified applicants.