Job Description

Minimum qualifications

• Master's degree in Finance, Economics, quantitative fields, or equivalent practical experience.
• 5 years of institutional finance experience in investigative, modeling-intensive roles (Quantitative Trading, Corporate Finance, or Alternative Asset Management).
• Experience with frontier LLMs, coding agents, or AI evaluation pipelines, including structured prompts and tool-use traces.
• Experience identifying flaws in AI-generated financial models, backtests, or memos, and translating these errors into generalizable grading rules.

Preferred qualifications

• CFA, CPA, MBA, JD, Master of Financial Engineering (MFE), or an advanced quantitative degree in Finance, Computer Science, Math, or Economics.
• Background building benchmarks, rubrics, or SFT/RLHF pipelines at AI labs, fintechs, or quantitative teams.
• Domain-specific excellence in Python and SQL (Quant) or advanced, macro-free spreadsheet software and Python/SQL (Alternative Asset Management/Corporate Finance).
• Familiarity with market data feeds, SEC EDGAR, Capital IQ, FactSet, Bloomberg, ERP/EPM, and tax/audit platforms.

Experience identifying flaws in AI-generated financial models, backtests, or memos, and translating these errors into generalizable grading rules.

About the job

DeepMind is building frontier financial reasoning and agentic capabilities across Gemini and our enterprise agent harnesses. Moving AI from a generalist research assistant to an agent that can autonomously complete financial workflows requiresmore than raw pre-training scale - it requires deep domain mechanics embedded directly into how models are post-trained, how agent harnesses and tools are architected, and how evaluations measure real-world finance.

As a Finance Subject Matter Experts (SME), you will work with models, evaluations, and post-training data in the following three Finance domains -

• Quantitative Trading
• Corporate Finance - specifically with experience across FP&A, Audit and Tax
• Alternative Asset Management

Unlike traditional advisory roles that stop at high-level feedback or static prompt writing, you will sit directly in the model development loop alongside Research Scientists, Post-Training Engineers, and Product Managers. You will prioritize differentiated financial workflows, design end-to-end financial benchmarks, construct atomic and deterministic grading rubrics, diagnose multi-step agent trajectories, repair near-miss model reasoning traces, design verifiable reward signals, and co-build domain-specific agent skills.

Artificial intelligence will be one of humanity’s most transformative inventions. At Google DeepMind, we are a pioneering AI lab with exceptional interdisciplinary teams focused on advancing AI development to solve complex global challenges and accelerate high-quality product innovation for billions of users. We use our technologies for widespread public benefit and scientific discovery, ensuring safety and ethics are always our highest priority.

We are pushing the boundaries across multiple domains. Our global teams offer diverse learning opportunities and varied career pathways for those driven to achieve exceptional results through collective effort.

Individual pay is determined by factors including job-related skills, experience, and relevant education or training.

US: $207000 - $300000 (USD) + 20% bonus target + equity + benefits

Learn more about benefits at Google.

We are pushing the boundaries across multiple domains. Our global teams offer diverse learning opportunities and varied career pathways for those driven to achieve exceptional results through collective effort.

Responsibilities

• Evaluate the AI's step-by-step process across calculations, ledger searches, and code to pinpoint quantitative or accounting errors.
• Create realistic financial tasks (e.g., valuations, backtests, audits) and write "gold-standard" examples to correct model near-misses.
• Build complex prompts featuring vague instructions, GAAP/non-GAAP differences, and conflicting estimates to test AI resilience.
• Create pass/fail checklists for deliverables and draft deceptive, polished-but-wrong answers to stress-test automated graders.
• Teach the AI to efficiently use core financial tools (SQL, etc) and categorize error patterns into reusable templates.