Applied AI Engineer

Greenboard•New York, NY

1d•$170,000 - $280,000•Onsite

About The Position

We're looking for an Applied AI Engineer to define and build the core AI systems powering Greenboard. This is an AI-first engineering role. You’ll work on model integration, agent behavior, and system design — while also owning how those capabilities translate into real product experiences. You’ll operate end-to-end: from designing how models should behave, to building the systems that support them, to shaping the user experience that makes them useful in a high-stakes, regulated environment. We’re a team that values truth-seeking, creativity, and strong opinions loosely held. We enjoy spirited debate, exploring unconventional ideas, and ultimately shipping. This is a full-time, on-site role in NYC.

Requirements

Strong experience working with LLMs or similar models — including prompt design, evaluation, and iteration
Experience with Typescript and building AI/ML powered systems in production
Ability to design systems around models (tooling, retrieval, orchestration), not just call APIs
Strong engineering fundamentals — comfortable building backend systems and APIs (frontend experience a plus)
Strong product instincts — you care about how systems behave in real user workflows
Ability to reason about tradeoffs between accuracy, latency, and cost
Clear communicator who can collaborate across technical and non-technical teams

Nice To Haves

Experience building agent-based systems, copilots, or AI-native products
Familiarity with retrieval-augmented generation (RAG), tool use, and evaluation frameworks
Experience working with structured data systems and building integrations across complex products
Background in fintech or other regulated industries

Responsibilities

Design and implement systems around LLMs and related models, including tool use, retrieval, memory, and orchestration
Own how models interact with internal systems and external data to produce reliable, grounded outputs
Develop and iterate on prompts, routing strategies, and context assembly to improve accuracy and usefulness
Define how the system chooses tools, retrieves information, and generates responses aligned with user intent and firm policy
Continuously improve system behavior so each release measurably increases quality and trust
Build evaluation frameworks to measure performance and detect regressions
Improve observability into model behavior and system outputs
Treat “wrong with confidence” as the primary failure mode — and design systems to minimize it
Translate AI capabilities into intuitive product experiences
Work across the stack to ensure tight integration between model behavior and user-facing features
Build pipelines that connect model outputs to sources of truth across the platform
Automate workflows by leveraging deep integrations with Greenboard’s underlying system of record
Partner with engineering, product, and domain experts to define problems and validate solutions
Work closely with customer-facing teams to ensure the system performs in real-world scenarios