Job Description
Job Description
Uber AI Solutions is Uber’s new marketplace connecting freelancers with Generative AI researchers. We’re inviting experienced finance professionals to collaborate on a new client project at the frontier of GenAI. This is a freelance, paid, project-based opportunity - flexible, remote, and designed for professionals who want to contribute their expertise while shaping the future of finance.
Responsibilities
As a Finance Domain Expert embedded within AI development, you will:
-
Workflow Definition: Define real-world, research-driven financial workflows (e.g., M&A analysis, valuation, portfolio construction) and translate them into structured LLM evaluation workflows.
-
Evaluation & Rubric Creation: Design rigorous evaluations, rubrics, and preference-ranking frameworks with clear inputs, expected outputs, and success criteria.
-
Model Capability Assessment: Evaluate LLM performance on complex financial tasks including financial modeling within LLM workflows, data extraction, statement analysis, market interpretation, and risk assessment.
-
Error & Quality Analysis: Conduct deep error analysis, identify failure modes, loss patterns, and quality trends, and feed insights directly into model improvement or fine-tuning cycles.
-
Data, Annotation & Feedback Loops: Guide evaluation dataset creation, annotation workflows, and human-in-the-loop processes (e.g., RLHF / preference optimization), working closely with product and engineering teams.
Project details
-
Location: Remote (You must be based in the United States)
-
Payout structure: Task-based pay model. Competitive rates per completed task, determined by the complexity and required experience.
-
Type: Freelance / Independent contractor.
Key Requirements
-
Proven Domain Expertise: 12+ years of hands-on experience in Investment Banking, Private Equity, Asset Management, or Equity Research.
-
Direct LLM Evaluation Experience (Non-Negotiable): Demonstrated experience in 3–4 or more of the following:
-
LLM projects (hands-on, not high-level strategy)
-
Model evaluation, preference ranking, or rubric design
-
Evaluation datasets and annotation workflows
-
Error analysis, failure modes, and quality trend tracking
-
Feedback loops into model improvement or fine-tuning
-
Human-in-the-loop / RLHF / preference optimization
-
Financial modeling embedded within LLM evaluations
-
Analytical Rigor: Ability to break complex financial reasoning into measurable, testable steps.
-
Communication: Can clearly explain nuanced financial judgments to non-finance and technical audiences.
-
Technical Acumen: Comfortable working closely with AI and engineering teams; understands how LLMs fail, not just how they work.
-
Attention to Detail: High bar for accuracy, consistency, and evaluation quality.
Why this matters
Your contributions could directly shape how Generative AI is applied in finance — from improving workflows to enhancing decision-making in investment banking, asset management, private equity, and equity research. This work has a global impact, helping create AI tools that meet the rigor and accuracy expected in the finance industry.