Agent Manager

Location: USA-Remote

Department: Product

About the Role

We are looking for a Product Manager for our AI agents. ShopMy is building a growing fleet of agents that do real work for internal teams and external clients, from brands and creators to the people who support them. As these agents move toward improving from their own track record, someone has to own what each agent is for and keep it pointed at useful work. That is you.

In this role you sit between the people an agent serves and the agentic system itself. You deeply understand the clients need, whether that client is an internal team or an external brand or creator, and you translate that need into concrete demands on the agents: how they should behave, how they delegate, what good output looks like, and where their boundaries sit. You work closely with the teams and clients each agent serves, along with product and engineering, turning recurring evidence from real runs into a system that gets measurably better over time.

What Were Looking For

We need someone who embodies three qualities:

You are a systems thinker. You deeply understand the people an agent serves, whether an internal team or an external brand or creator, and you see how their work actually flows: where decisions happen, where context gets lost, where things slow down. You assume the requests and complaints you are handed are incomplete, and you read between them to find the need actually worth solving. You know where an agent genuinely helps and where it does not.

You are hands-on with agents. You do not just write requirements and hand them off. You are in the prompts, documents, and delegation yourself, you prototype and test against real runs rather than writing long specs, and you partner with engineering to ship the harder capabilities. You earn trust through what works.

You live on the frontier of agents. You have real intuition for LLMs: how they are prompted, how they delegate to one another, and how they are evaluated. You set the bar for what a great agent feels like, accurate, reliable, on-tone, and trusted, and you keep up with what is newly possible.

What You’ll Do

• Deep dive with customer teams to deeply understand their processes and problems, and decide and prioritize where to plug in agents for the greatest impact
• Own and deeply understand the customer need behind each agent, whether the client is an internal team or an external brand or creator, and convert it into clear requirements and priorities for the agents
• Garden agents against that need: watch for recurring gaps, quality drift, confusing delegation, weak eval coverage, and agents being stretched beyond their shape
• Own prompt and document quality, delegation shape, locked instructions, and agent boundaries for your agents with a high bar for precision
• Help shape the agentic orchestration system itself, not just the individual agents: spot flaws, friction, and missing capabilities in how agents are coordinated, and act as a product driver for where the system should go next
• Triage the improvement queue: act on the agent-side changes you own, route tool and capability gaps to engineering, and route product ambiguity to product owners
• Partner with the people an agent serves, along with product and engineering, so every agent has the right scope, knowledge, and tools to succeed, without taking over their domains

What Were Looking For

• 3+ years in product management, or a closely related role where you owned a product, workflow, or service against real customer needs
• A keen understanding of business processes and strong business judgment: you can reason about where the leverage and the ROI are, and focus the agents there
• A track record of deeply understanding customers, internal or external, and turning their needs into clear requirements others can act on
• An evidence-driven, analytical mindset, with the patience and attention to detail to read run histories, spot recurring patterns, and prioritize by impact
• Enough technical fluency to prototype, evaluate, and go deep with engineers on agent design and reliability; you do not need to be a software engineer, but you cannot be hands-off
• Real comfort operating independently amid ambiguity, in a role and a system that are both still being built
• Bachelors degree or equivalent practical experience

Bonus Points

• Experience working in high-growth startups or fast-paced product environments
• Hands-on experience building, prompting, or evaluating LLM agents
• Familiarity with eval frameworks or LLM-as-judge systems
• Background in consulting or strategic business operations
• Background in the creator economy, influencer or affiliate marketing, or e-commerce
• Experience working directly with brands or creators as clients