7 hours ago
New York, NY, USAMid Level
Responsibilities
- Own AI tooling adoption by identifying high-leverage opportunities, building or integrating solutions, teaching users, and driving sustained adoption.
- Build internal AI services, workflows, interfaces, model integrations, and agent infrastructure.
- Improve agent performance and trust through codebase context, tool access, review, testing, CI, permissions, secrets, and safe rollout.
- Partner with operations, deployment, data, and business teams to build maintainable workflow systems and hand them off to clear owners.
- Create reusable connectors, shared context, templates, examples, and guardrails for other teams.
- Evaluate whether AI tools should be adopted, composed, or built internally.
- Partner with infrastructure and security teams to provide useful and safe access to AI tools.
- Measure adoption, productivity, and user trust, expanding effective tools and discontinuing ineffective ones.
Requirements
- Strong software engineering fundamentals and experience shipping backend services, internal tools, integrations, automation, and interfaces using Python and/or TypeScript.
- Experience shipping agentic systems or LLM-powered tools used by real people, including understanding their failure modes.
- Fluency with modern AI coding tools and sound judgment about when to build, buy, use agents, or use conventional scripts.
- Understanding that evaluations, permissions, secrets, and safe rollout are part of the product.
- Ability to work with non-engineering teams, scope ambiguous workflows, communicate clearly, and drive adoption.
- At least 4 years of production software experience is a reasonable bar, with demonstrated delivery of comparable systems valued more than the precise year count.
- Preferred qualifications include startup founder experience, self-improving agentic systems that operated without continuous human involvement, strong product sense, UX experience in human-in-the-loop systems, and measured business workflow redesign experience.
Benefits
- In-person role based in New York City.
- Direct exposure to how Ultra builds and operates.
- Opportunity to work across engineering, infrastructure, developer productivity, operations, deployment, data, people, and business teams.
Tech Stack
Categories
AI ApplicationsProduct Engineering
About Ultra
Ultra builds AI-powered software for customs agencies and trade professionals to streamline compliance workflows, manage risk, and process shipments accurately. Its products include Publican for customs authorities and Publican World for brokers and import/export teams, sold as enterprise SaaS. Founded in 2016 and headquartered in Jersey City, NJ, the company is privately held.