Finance AI Data Trainer - Phase 3, Findex

Finstock, Inc. builds AI-powered financial research, trading analytics, quantitative research, and market intelligence tools for experienced market users, analysts, research teams, and institutions. Our product, Findex, is a market intelligence workspace designed to help users explore financial data, market views, charts, research prompts, and workflow-oriented questions.

About The Role

We are hiring Finance AI Data Trainers - Phase 3 to help improve Findex's local workspace agent. In this role, you will evaluate, annotate, and improve AI-generated financial research outputs. You will help Findex better understand finance-related user questions, identify missing context, structure market research responses, flag unsupported claims, and produce clearer research-support answers.

This role is ideal for candidates with strong finance knowledge, analytical writing skills, attention to detail, and an interest in AI model evaluation, financial data quality, and responsible AI for market research workflows. This role does not involve managing client funds, executing trades, providing personalized investment advice, or guaranteeing financial outcomes.

Responsibilities

• Review AI-generated responses to finance, investment research, market analysis, and risk-related prompts.
• Evaluate model outputs for financial accuracy, reasoning quality, clarity, evidence use, data limitations, and appropriate risk framing.
• Create, annotate, and improve finance-related prompts, expected answers, scoring rubrics, and edge-case examples.
• Identify hallucinations, unsupported assumptions, misleading financial claims, calculation errors, and weak reasoning.
• Rewrite or improve AI responses so they are clearer, more accurate, more structured, and aligned with responsible research-support standards.
• Test Findex workflows involving market data, charts, research prompts, financial statements, macro context, portfolio/risk context, and user-scoped workspace information.
• Help define when the AI should answer directly, ask clarifying questions, disclose limitations, or avoid making personalized recommendations.

Requirements

• Strong understanding of finance, financial markets, investment research, accounting, economics, or quantitative analysis.
• Ability to evaluate financial reasoning across topics such as equities, valuation, financial statements, macroeconomics, risk analysis, portfolio concepts, FX, crypto assets, options, or trading research.
• Excellent written English and ability to explain complex financial issues clearly and objectively.
• Strong attention to detail and ability to detect subtle errors in logic, assumptions, calculations, and interpretation.

Qualifications

• Experience in AI data annotation, RLHF, model evaluation, prompt evaluation, financial QA, or data quality review.
• Prior experience in investment research, equity research, financial analysis, trading analytics, portfolio analysis, or market intelligence.
• Progress toward or completion of finance-related credentials such as CFA, FRM, CPA, CAIA, or equivalent professional training.

What You Will Work On

• Comparing multiple AI-generated financial responses and selecting the stronger answer based on a scoring rubric.
• Rewriting weak model outputs into clearer, more factual, and better-structured research-support responses.
• Creating finance prompts that test Findex's ability to handle realistic user questions.

Important Role Boundaries

This position supports product quality, AI evaluation, financial data annotation, and research-workflow improvement. The role does not require or permit the contractor to:

• Provide personalized investment, legal, tax, accounting, or financial advice to users.
• Recommend that users buy, sell, or hold any financial instrument based on their personal circumstances.
• Execute trades or manage funds.
• Handle client assets, deposits, or payments.
• Promise or imply investment returns, trading profits, or risk-free outcomes.

Pay

USD $35-$75 per hour, depending on demonstrated finance expertise, AI evaluation experience, task complexity, and work quality.

Schedule

Approximately 20 or 40 hours per week for a 16-week contract, with possible extension based on project needs and performance.

Remote - United States preferred; candidates in other countries may be considered where Finstock, Inc. is able to engage contractors in compliance with applicable laws and project requirements.