Senior Business Analyst & Prompt Engineer (AI Skills)

Job Description

Job-Senior Business Analyst & Prompt Engineer (AI Skills)

\n

Job Location - NYC, NY

\n

Job Type- Full Time

\n

\n

The role

\n

A senior, independent contributor who designs, tests and maintains reusable AI "Skills": structured prompts, instructions and evaluation sets that let AI agents do work across the product development lifecycle. Examples include drafting business intent documents, generating epics and stories, and checking requirements against risk and policy standards. You'll work directly with senior product, technology and risk stakeholders. This is not an engineering role: we want someone who knows how models behave and can reliably get the results the business needs.

\n

\n

Location: New York City, hybrid (3 days a week onsite)

\n

Engagement: Embedded with a technology and operations team at a large global bank

\n

Level: Senior, independent contributor

\n

\n

What you'll do

\n

\n
• Turn business processes (ideation, business case, requirements, design review, testing) into well-defined AI Skills with clear inputs, outputs, quality criteria and escalation points.
\n
• Write, test and refine prompts and instruction sets for large language models, using structured outputs, few-shot examples, role and context framing, and step decomposition.
\n
• Build evaluation sets and rubrics that measure completeness, consistency and accuracy of Skill outputs, and iterate on failures until quality is production-ready.
\n
• Run discovery with senior stakeholders to define what "good" looks like, and document it as business requirements, epics and user stories with acceptance criteria.
\n
• Define where humans approve and where agents proceed on their own, including approval gates, review points and exception handling.
\n
• Work with risk, compliance and control teams so Skills reflect policy, regulatory and audit expectations from the start.
\n
• Maintain a versioned library of Skills with documentation, usage guidance and change history.
\n
• Compare models and model versions on specific tasks, and recommend which suits which job, flagging cost, latency and reliability trade-offs.
\n
• Coach product and business teams on effective use of AI tools.
\n

\n

\n

What we're looking for

\n

Must have

\n

\n
• 8+ years in business analysis, product, process or consulting roles, including 2+ years of hands-on prompt engineering or LLM-based solution design (assumed range; adjust as needed).
\n
• Deep working knowledge of how leading LLMs behave (strengths, failure modes, hallucination, consistency, instruction-following) and how to steer them.
\n
• Strong requirements skills: BRDs, epics, user stories, Gherkin or similar acceptance criteria, process and workflow mapping.
\n
• Proven ability to work independently, influence senior stakeholders, and write clearly.
\n
• Experience in financial services or another regulated industry, and comfort in risk, compliance and control environments.
\n
• A learning mindset: able to describe what you're proud of, what failed and what you learned, and open to being coached and adapting your approach.
\n

\n

\n

Nice to have

\n

\n
• Experience with Gemini or other enterprise LLM platforms, and AI-assisted development tools such as GitHub Copilot/Devin/Kiro/Codex/Claude Code/Cursor
\n
• Familiarity with Jira, Confluence and Figma, and with agent or multi-agent workflow concepts.
\n
• Experience designing LLM evaluation, red-teaming or "LLM-as-judge" validation approaches.
\n
• Exposure to the software or product development lifecycle from idea to release.
\n

\n

\n

Not required

\n

\n
• Software engineering depth. Light Python or API familiarity helps but is not expected.
\n

\n

\n

Success looks like

\n

\n
• A tested library of Skills in use across multiple stages of the lifecycle.
\n
• Measurable improvement in output quality, consistency and cycle time against agreed evaluation criteria.
\n
• Stakeholders who trust the Skills and can adopt them without hand-holding.
\n

\n