Product Engineer, AI & Automation

LINCOLN CENTER•New York, NY

1d•$120,000 - $150,000

About The Position

The Operational Efficiency team works with departments across the organization to reduce manual effort in daily operations. This role is the team's primary builder. You will take processes that are currently slow, repetitive, or error-prone and turn them into working automations, applying AI capabilities where they fit the problem. Most of your time will be spent building: writing code, connecting systems, designing and testing agents, and maintaining what you have already shipped. Requirements will reach you as problems rather than finished specifications, so you will work directly with the teams whose processes you are automating to understand how the work is performed before deciding what to build. We are looking for someone with five or more years of experience who has delivered automations that people depend on daily and who has supported them over time.

Requirements

5+ years of professional experience building automations, integrations, or internal tools.
Strong Python, JavaScript, or TypeScript skills, with experience writing code that other engineers have maintained.
Substantial experience with REST APIs, authentication patterns, webhooks, and system-to-system integration.
Hands-on experience building with large language models, including prompt design, structured output, and methods for evaluating output quality.
Working knowledge of SQL and experience with inconsistent or incomplete operational data.
Proven ability to work with non-technical stakeholders, including gathering requirements, managing expectations, and supporting adoption.
Good judgment about which processes are appropriate for candidates for automation, and which are not.

Nice To Haves

Experience with integration or workflow platforms such as Workato, Make, n8n, Zapier, Power Automate, Airflow, or Temporal.
Experience with document processing, OCR, or extraction from unstructured sources.
Experience in a comparable operational environment where automations touch finance, HR, or customer-facing workflows.
Familiarity with data governance and access controls in environments that handle sensitive records.

Responsibilities

Design and build automations across operational workflows, including data movement, document handling, reporting, intake and routing, and notifications.
Connect existing systems using REST APIs, webhooks, scheduled jobs, and integration platforms, working within the organization's current technology stack wherever possible.
Build components that use large language models where they are the right tool for the problem, including extraction from unstructured documents, classification, routing, summarization, and drafting. Design the prompts, test them against representative data, and define the accuracy required before launching.
Determine where human review is required, set confidence thresholds that route uncertain cases to staff, and define what happens when the model produces incorrect output.
Handle supporting work with the same care as the build itself, including data validation, idempotency, credential management, environment separation, and documentation sufficient for another engineer to take over.
Work with department leads to document current-state processes and identify automation candidates, quantifying the time and cost each process consumes today.
Assess feasibility and recommend alternatives when a process change or a configuration change in an existing system would solve the problem more effectively than a new automation.
Translate operational needs into technical requirements and divide work into phases that deliver value early.
Set clear expectations with stakeholders on accuracy, exception handling, and scope. This matters especially for AI-assisted steps, where acceptable error rates should be agreed in advance.
Own each automation through its full lifecycle, including monitoring, response when upstream systems change, iteration as processes evolve, and retirement once it is no longer needed.
Build in monitoring and alerting so that failures surface quickly instead of going unnoticed.
Maintain documentation and runbooks so that support does not depend on a single person.
Track and report impact, including hours returned, error rates, cycle time, and usage.
Train and support staff who want to build simple automations on their own, extending the reach of the team.
Contribute to the team's standards and patterns, including guidelines on what organizational data may be sent to external AI services.