What are the responsibilities and job description for the Technical Product Owner position at Sholder?
Technical Product Owner
Build a six-month product roadmap with an AI-native team — then keep it rolling.
Position Details
This is part-time, contractor position. The work we have will take about 8-12 months. You can view compensation details at https://jobs.sholder.com.
The short version
Sholder's product roadmap is currently at zero months out. We need someone to build it from scratch — six months of prioritized, sequenced, well-reasoned product direction — working directly alongside Jarvis, our AI development system. If that goes well, we'd like you to stay on a light retainer to keep the horizon rolling.
About Sholder
Sholder is building the infrastructure for human support — a platform that connects people with their own network of trusted advisors, coaches, and peers, powered by an AI layer that makes every relationship more intentional and every session more effective.
We're a small, focused team. Our product is live. Our users are real. And our development model is unusual: the majority of our engineering is driven by an AI system named Jarvis, working within a codebase and product framework built to move fast without losing quality.
How this engagement works
This is a two-phase engagement, and we want to be upfront about that. Phase 1 is an hourly contract to build the initial 6-month product roadmap in collaboration with Jarvis. You work the hours it takes to do it well — getting deep context on where the product is, where leadership wants it to go, and what the AI development system is capable of executing. The output is a prioritized, sequenced plan the team can actually build against.
Phase 2 is an optional retainer. If the project goes well and you want to stay involved, we'd like to keep you on to roll the roadmap forward — reviewing what shipped, adjusting for what we learned, and maintaining a 6-month horizon on a lightweight monthly cadence. This is roughly 10–20 hours per month, async-first.
We chose this model deliberately. An hourly Phase 1 lets both sides verify alignment before committing to something longer — you're not locked into a fixed scope, and we're not locked into a fixed price. If Phase 1 delivers a roadmap we trust and a working relationship that clicks, Phase 2 is the obvious next step. If it doesn't, both sides walk away cleanly.
What this role is, honestly
This isn't a traditional TPO role. You won't be managing a team of engineers in daily standups or grooming a backlog in Jira. You'll be working as the human judgment layer in an AI-accelerated build cycle — defining what to build, reviewing what gets built, and making sure the two stay aligned.
Your primary development partner is Jarvis, Sholder's AI development system. That means your job isn't to translate requirements into engineering tickets — it's to translate business and user needs into context, constraints, and acceptance criteria precise enough that an AI can act on them correctly and a human can verify it did.
What you'll doPHASE 1 — ROADMAP BUILD (WEEKS 1–6)
- Get deep context on the product: current state, user signals, leadership vision, and what Jarvis is capable of building
- Define and sequence a 6-month product roadmap — prioritized, realistic, and grounded in what will move the business
- Collaborate directly with Jarvis to validate feasibility and refine the plan against real execution constraints
- Produce the process artifacts that make the roadmap maintainable: intake framework, prioritization criteria, AI-agent collaboration playbook
- Align leadership on the roadmap and get buy-in on the sequencing decisions
PHASE 2 — RETAINER (ONGOING, MONTHLY)
- Review what shipped each month against plan — surface drift, validate quality, update the roadmap accordingly
- Roll the 6-month horizon forward: intake new signals, re-prioritize, add net-new features as the product learns
- Build and maintain feedback loops: user interviews, support signal synthesis, stakeholder reviews
- Represent the product to partners, share members, and service providers as needed
- Track quality metrics specific to AI-assisted platforms: accuracy, behavioral consistency, trust signal trends
Who you are
You've owned a technical product backlog before — not just contributed to one, but owned it. You know the difference between a user story and a requirement, and you can write both with the precision that stops engineering debates before they start.
You're comfortable with ambiguity and fast-moving contexts. You know when to push for clarity and when to make a call and move forward. You have enough technical grounding to have a real conversation about system design, API contracts, and AI/LLM behavior — without needing to be the one who implements any of it.
Ideally you've done this kind of engagement before — come in as a fractional or contract PM, delivered something concrete, and either wrapped cleanly or transitioned to an ongoing relationship. You know how to be effective without being embedded full-time.
YOU BRING:
- 5 years in product ownership, product management, or a technical BA/PM hybrid role
- Experience with AI-native products or LLM-backed features in production — not just experimentation
- A track record of building or rebuilding a product roadmap from low structure, not just maintaining one that already exists
- Data fluency: comfortable with SQL, dashboards, and defining the metrics that distinguish signal from noise
- Strong communication across technical and non-technical audiences
- Prior fractional, contract, or embedded product leadership — you know how to deliver without full-time context
NICE TO HAVE:
- Experience with prompt engineering, system prompt governance, or AI evals
- Familiarity with React/TypeScript and Go — enough to read code and understand what's been built
- Background in consumer or B2B2C SaaS at seed to Series A stage