doss·San Francisco

Agent Engineer

On-siteAshbyDirect apply link

About Doss

DOSS is building an Operations Cloud for the real world. We deliver a modern, AI-native platform that helps physical product businesses manage the flow of goods, dollars, and data across procurement, inventory, orders, fulfillment, and finance in real time. Purpose-built to replace spreadsheet chaos and rigid, consultant-heavy ERP implementations, DOSS delivers fast time-to-value and stays adaptable as a business evolves. Through our Adaptive Resource Platform (ARP) and unified operational data model, teams can deploy quickly, automate workflows, and make changes without months of re-implementation. We recently raised a $55M Series B co-led by Madrona and Premji Invest, with participation from Intuit Ventures, Theory Ventures, General Catalyst, Contrary Capital, and Pathlight VC. DOSS is trusted by fast-growing operators to run critical operations with speed, control, and confidence.

About the Role

This role sits inside the product + implementation org (PMO), but is an engineering function: you’ll build the tooling and productize agents that operate across the DOSS customer lifecycle (discovery → design → build → UAT → go-live → expansion). You won’t be “writing code” in the classic feature-ticket SWE loop. Instead, you’ll be doing systems engineering for agentic software: designing context, interfaces, instrumentation, evaluation, and reliability so agents can execute real customer work safely and repeatedly.

Our goal is to turn ERP implementation into a mechanically scalable factory; breaking down knowledge work into computable units of context that can be run by GPUs, not human brains. You are going to help architect, instrument, and operate this factory.

While having previous ERP or industry experience is always a plus, it is not a pre-requisite. Our current team comes from a mix of business and technical backgrounds. What we value the most are:

What You'll Do

Key Responsibilities:

Architect agent workflows that map to real implementation work: intake → reasoning → actions → verification → handoff, across many customers and industries.
Design “context systems”: what an agent should know, when it should know it, how it’s grounded, and how it stays correct as the customer evolves.
Productize agent capabilities into reusable primitives: playbooks, tools, schemas, guardrails, and templates that compound across customers.
Instrument and evaluate: define success metrics, build automated eval harnesses, track failure modes, and iterate rapidly.
Orchestrate agents in production: reliability, rollback strategies, escalation paths, auditability, and continuous improvement loops with the PMO team.
Partner deeply with PMO + product to convert messy business reality into clean, computable representations (data models, process graphs, integration surfaces).
Drive automation strategy inside implementations: identify the highest-leverage “knowledge work → context unit” cuts and turn them into shipped systems.

What We're Looking For

2-10 years in engineering (industrial, process, computer etc.), consulting, finance, operations, product operations or technical program management at a technology company - neither Software Engineering or ERP experience not required (though if you have experience in that its a plus)
Systems Thinker. Everything in our business is highly interconnected across people, process, and technology. In order to make progress you need to be able to traverse it, untangle it, and put it back together.
High-Intensity. You have experience in high-pressure roles within startups, consulting, finance etc. - these are your real preferences, not your stated preferences.
Technical from 1st Principles. You can grok hard engineering and logic problems from a place of intuition, not memorization.
Intellectual Firepower. You cut through noise, synthesize, and get to the right answers fast.
Strong analytical mindset; experience with SQL is a plus (but not required)