This is a role posted by Reval Recruiting on behalf of a client

Software Engineer

Location: New York, NY

This role is posted on behalf of the following client: A fast-growing applied AI company building agents that automate operational supply chain workflows for businesses. The team partners directly with customers to deploy agents into production systems (including ERPs) and are focused on becoming a category leader at the intersection of AI and operational supply chains.

Role Overview

Join a small, high-caliber engineering team building and deploying AI agents inside real businesses. This role focuses on architecting multi-step agent systems, integrating with live operational/ERP data, and shipping reliable agents that drive measurable ROI in production. The work blends engineering, systems thinking, and product building in ambiguous, high-stakes environments.

What You'll Do

• Build and ship AI agents that automate operational workflows (e.g., purchasing, inventory management, order intake, customer support).
• Architect multi-step agent systems with tool use, structured outputs, memory, and orchestration across complex workflows.
• Develop context and memory systems so agents can reason over transaction history, customer behavior, and SKU-level operational data.
• Work directly with live ERP and ops data that is incomplete, inconsistent, and production-real.
• Create evaluation, monitoring, and feedback loops to continuously improve agent reliability over time.
• Partner directly with customers in a forward-deployed model—mapping ambiguous processes and turning them into AI-native workflows.
• Help define the long-term agent architecture and platform layer that future engineers will build on.

Who You Are

• Strong engineering fundamentals, sharp critical thinking, and the ability to clearly defend technical perspectives.
• Experience building with LLMs in production (beyond prototypes), including familiarity with agent frameworks, tool use, orchestration, and multi-step reasoning systems.
• Comfortable working with messy, unstructured, real-world operational data and translating ambiguity into robust system behavior.
• Product-minded with strong instincts for usefulness and real-world outcomes (not just benchmarks).
• High-ownership builder who thrives in fast iteration cycles and minimal handholding.
• Bonus: experience in operational/workflow-heavy domains (e.g., supply chain, logistics) and/or prior startup experience.