Software Engineer (AI Agents)
$140,000–$200,000 year
On-siteNew York City, New York, United States
Job Summary
Embed with customers and operators to understand supply chain workflows, then design and ship production agent harnesses on frontier LLMs that automate meaningful tasks. Build agent systems handling tool use, sub-agents, retrieval, and orchestration while architecting data, services, and APIs that integrate with customers' WMS, TMS, and ERP environments. Own evaluation as a first-class discipline by creating datasets, rubrics, and graders from real traces to prove reliability and unit economics. Codify repeatable deployment patterns to accelerate customer rollouts. This role is a founding position within Traba's Applied AI team, backed by Founders Fund, Khosla Ventures, and General Catalyst, focusing on transforming global supply chain operations through autonomous agents.
Required Qualifications
- 1+ years of software engineering experience
- track record of shipping LLM- or agent-based features to production
- Strong in Python and/or TypeScript/Node.js
- comfortable designing APIs, distributed systems, and data models in PostgreSQL
- Hands-on with the modern agent stack: production-scale prompt engineering, evaluation frameworks, orchestration patterns, and frontier model APIs
- A track record of building in fast, ambiguous environments—ideally at a vertical AI, AI-agent, forward-deployed, or data-product company
- Excellent written and verbal communication
Desired Qualifications
- Builder with an AI operator's instinct
- Domain-immersed
- High-agency in ambiguity
- Sweat both ends of the stack
- You've shipped real product on top of LLMs—not just chat wrappers
- You've designed agent harnesses, structured tools, written evals, and tuned prompts against production traces
- You think in capability, reliability, and unit economics—not just whether the model says the right thing
- You enjoy time with the people who do the work—learning an industry's vocabulary, edge cases, and operational tempo—and let that shape what you build
- Dropped into a fuzzy customer problem with a half-formed hypothesis and a deadline, you scope, build, evaluate, and ship without waiting for a spec
- You move between prompt iteration, eval design, backend services, and customer-facing UI in the same week
- You care about evals that catch regressions and traces that are easy to debug
- Embed with customers and operators to understand how supply chains run today—then design and ship agents that take meaningful work off their plate
- Build production agent systems on frontier LLMs: tool use, sub-agents, retrieval, structured outputs, MCP servers, and the orchestration that ties them together
- Own evaluation as a first-class discipline—datasets from real traces, rubrics and graders, experiments, and improvements you can prove move the needle
- Architect the data, services, and APIs the agent layer depends on—integrating our internal systems with customers' WMS, TMS, and ERP environments
- Codify repeatable deployment patterns so each new customer rollout is faster than the last
- ideally at a vertical AI, AI-agent, forward-deployed, or data-product company
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.