Software Engineer, AI Platform

UMATR; New York (NY); On-site; USD 140,000 - 190,000; Full time

Job summary

UMATR is a high-growth technology company building software and AI solutions for teams working with complex physical systems. Their platform handles telemetry, time-series data, test results and engineering logs, and they’re building AI systems to reason over this data to help engineers investigate failures and compare test runs.

They are looking for a strong software engineer with hands-on experience building production AI and agent systems to take ideas from 0→1 through to production.

Qualifications

• 5+ years of software engineering experience.
• Meaningful hands-on experience building AI/agent systems that operate in production.
• Strong software engineering ability; engineering-first role, not purely ML research.
• Experience with agent infrastructure, LLM applications, retrieval/context systems, evaluation or AI reliability.
• Comfortable with large-scale data, APIs and distributed systems.
• Strong product instincts and ability to work directly with users or customers.
• Comfortable making technical decisions independently in a 0→1 environment.
• Track record of owning systems beyond simply integrating an LLM API into an existing product.

Responsibilities

• Build the infrastructure and tooling behind production AI agents and agentic workflows.
• Develop systems for retrieval, context management, evaluation and AI observability.
• Build reliable backend and distributed systems that support AI workloads at scale.
• Design and improve evaluation frameworks for measuring agent quality, reliability and performance.
• Work with large volumes of telemetry, time-series data, test results and engineering logs.
• Prototype new AI capabilities quickly, test with users and productionise what works.
• Work directly with engineers and customers to understand real-world problems and turn them into technical solutions.
• Make architectural decisions in a 0→1 environment, balancing speed with long-term reliability.
• Own projects end-to-end, from initial idea and prototype through to production.

Skills

Backend systems

Distributed systems

AI/agent systems

LLM applications

Retrieval & context

Evaluation & AI reliability

APIs & data-scale

Independent decision making

Product instincts

End-to-end ownership

Job description

We’re partnering with a high-growth technology company building software and AI solutions for teams working with complex physical systems.

Their platform already handles large volumes of telemetry, time-series data, test results and engineering logs. The next step is building AI systems that can reason over this data and help engineers investigate failures, interrogate telemetry, compare test runs, identify anomalies and better understand how physical systems are behaving.

They’re looking for a strong software engineer with hands-on experience building production AI and agent systems to help take this from 0→1.

What You’ll Do

• Build the infrastructure and tooling behind production AI agents and agentic workflows.
• Develop systems for retrieval, context management, evaluation and AI observability.
• Build reliable backend and distributed systems that support AI workloads at scale.
• Design and improve evaluation frameworks for measuring agent quality, reliability and performance.
• Work with large volumes of telemetry, time-series data, test results and engineering logs.
• Prototype new AI capabilities quickly, test them with users and productionise what works.
• Work directly with engineers and customers to understand real-world problems and turn them into technical solutions.
• Make architectural decisions in an ambiguous 0→1 environment, balancing speed with long-term reliability.
• Own projects end-to-end, from initial idea and prototype through to production.

What We’re Looking For

• 5+ years of software engineering experience, with strong backend/distributed systems fundamentals.
• Meaningful hands-on experience building AI/agent systems that operate in production.
• Strong software engineering ability - this is an engineering-first role rather than a research-focused ML position.
• Experience with agent infrastructure, LLM applications, retrieval/context systems, evaluation or AI reliability.
• Comfortable working with large-scale data, APIs and distributed systems.
• Strong product instincts and the ability to work directly with users or customers.
• Comfortable making technical decisions independently in a genuinely 0→1 environment.
• A track record of owning systems beyond simply integrating an LLM API into an existing product.