Staff Applied AI Engineer — New York, NY

About the opportunity

Vx5FPv ju5sC2 is partnering with a well-established, venture-backed B2B SaaS company with more than a decade of product usage data across thousands of software teams. They’re building a new applied AI team from scratch to ship AI-native product intelligence: systems that connect what users do inside software with the code behind it, then recommend improvements with a human in the loop.

This is not an AI layer bolted onto an existing product. The company’s whole platform is moving onto it. The team moves fast, has high autonomy and builds things without clear precedents. Applied AI engineers sit on product teams alongside the company’s core software engineers and are responsible for getting AI features in front of enterprise customers.

This is a Staff-level role. You’ll help set technical direction for how AI gets built across the product, guide the engineers you’re paired with and own outcomes, not just code.

What you’ll do

• Design and build applied AI systems, including RAG pipelines, agentic workflows and LLM-powered features, from prototype through production

• Help set the patterns other engineers follow for building with LLMs: architecture, evaluation, guardrails and code review

• Decide when to prompt, when to fine-tune and when to use a different tool entirely, and explain those tradeoffs to the team

• Build evaluation frameworks, catch hallucinations early and implement guardrails that hold up under real load

• Own MLOps end to end: deployment, monitoring, latency, cost and reliability

• Build human-in-the-loop workflows that keep people in control of AI recommendations

• Contribute full-stack when the product needs it

• Partner with product and design to frame problems before writing code, and push back when the framing is wrong

What you bring (required)

• 7+ years of professional software engineering, with a product engineering focus. We don’t expect 7 years of AI experience; almost no one has that. We want a strong engineering and product background with recent, hands-on LLM work on top of it.

• Experience at a large, established tech company

• Tech lead experience: setting technical direction, mentoring engineers and owning delivery on a product team

• LLM-powered features you personally built and shipped to customers in production, including RAG, tool use or agent orchestration. Prototypes, hackathons and internal proofs of concept don’t count on their own.

• Strong command of model evaluation and systematic hallucination mitigation

• Production MLOps experience in a live environment

• Strong full-stack fundamentals

• Clear communication with engineers, product managers and executives

• Product judgment: you ask “should we build this?” before “how do we build this?”

Nice to have

• You took an AI product to market and scaled it to enterprise customers

• You moved from a large tech company to a founding or early engineering role at an AI startup

• Fine-tuning experience and a point of view on when it beats prompting

• Experience with MCP, LangChain, LlamaIndex or custom orchestration

• A background in SaaS or product analytics

This role is not a fit if

• You have fewer than 7 years of professional software engineering experience

• Your AI work is limited to prototypes, coursework, research or internal proofs of concept

• You haven’t led engineers or owned technical direction on a product team

• You don’t currently live in the NYC metro area

• You need visa sponsorship now or in the future

Compensation and benefits

• $255,000 to $276,000 base, plus equity. Final offer depends on experience. Vx5FPv ju5sC2 negotiates on the candidate’s behalf.

• Employer-heavy health coverage with $0 premium options

• Strong 401(k) match

• Flexible time off

Logistics

• Must currently live in the NYC metro area. On-site in Manhattan 3 days a week, including two set team days.

• Must be authorized to work in the US without current or future visa sponsorship

• No relocation assistance

How to apply

Submit your resume with a short note (a few sentences is fine) that answers:

• One LLM-powered feature you shipped to customers: what it did, who used it and at what scale

• How you evaluated it and how you handled failures in production

• What you’d do differently if you built it again

Applications without the note won’t be reviewed.

About Vx5FPv ju5sC2

Vx5FPv ju5sC2 is a cooperative talent network. This search is represented by Vx5FPv ju5sC2. The client is an equal opportunity employer and welcomes people of all backgrounds, experiences, abilities and perspectives.

• Questions? Reach out to Jamar at •••••••• or (••••••••, or connect at ••••••••