Venture Launchpad Product Manager (AI-Native Product Operations)

# Venture Launchpad Product Manager (AI-Native Product Operations) **Status:** Draft for review **Location:** Remote; meaningful overlap with US working hours preferred **Engagement:** [Full-time / contract-to-hire] **Reports to:** Founder **Works closely with:** Engineering lead, QA lead, product owners, legal/compliance advisers, and internal AI agents ## The role We are building a repeatable launchpad for turning a portfolio of technically ambitious products into validated, launch-ready businesses. The products may involve AI, developer infrastructure, games, on-chain systems, prediction markets, data products, or new forms of digital coordination. We need a Product Manager who can build and operate the machinery behind those launches—not merely maintain separate roadmaps after the fact. Your job is to take a product opportunity from ambiguous founder intent to a bounded launch thesis, an executable plan, a qualified release, and a measured market result. You will create the reusable operating system that makes the next launch faster and more reliable than the previous one: intake criteria, decision records, launch templates, dependency maps, QA gates, evidence requirements, distribution plans, retrospectives, and portfolio-level visibility. An early focus will be applying the launchpad to one technically ambitious product and then proving that the operating model transfers to a second. Some portfolio opportunities may involve on-chain deployments, prediction mechanisms, or token-related capabilities, but every such initiative is subject to explicit technical, economic, security, legal, and jurisdictional gates. Moving quickly does not mean bypassing those gates. This is not a conventional project-coordinator role. We are looking for a reasonably technical product operator who can reason across markets, product behavior, APIs, data models, authentication, smart contracts where relevant, infrastructure, launch economics, quality, and distribution—and who can direct humans and AI agents toward a verified outcome. An operations-style product person can succeed here, but only if they can engage directly with technical artifacts and do not need every engineering question translated for them. ## How we work: humans directing agents We are building an internal execution and coordination system comprising persistent AI agents, shared project state, and auditable work handoffs. The system can accept work, route tasks to the appropriate agent or human, maintain project context, track dependencies, return artifacts for review, and preserve a record of decisions and outcomes. You will be expected to operate at **AI velocity**. That means using agents for research, opportunity analysis, competitive mapping, requirements synthesis, technical discovery, planning, ticket preparation, QA design, documentation, launch collateral, metrics analysis, follow-up, and portfolio reporting. You will have an agent identity within our system and may direct multiple specialist agents or project-specific cells. Your responsibility is not to produce the most agent activity. It is to produce trustworthy outcomes: the right work, with the right context, inside the right authority boundary, supported by evidence another person can inspect. You may bring a strong existing agent stack—commercial agents, open-source systems, custom agents, workflow automation, or other tools—provided it can interface with shared project state and does not create an opaque parallel company. We expect you to understand context, memory, task decomposition, concurrent execution, verification, escalation, and recovery as management concerns, not merely prompting techniques. AI velocity does not remove human accountability. You remain responsible for product judgment, factual accuracy, prioritization, permissions, quality, and the final recommendation. ## What you will own ### Build the launchpad operating model - Define how opportunities enter the portfolio, what evidence advances them, and what conditions stop or defer them. - Create reusable launch primitives: opportunity brief, customer and market thesis, decision log, dependency graph, risk register, release criteria, QA matrix, launch plan, metrics plan, and retrospective. - Establish a visible sequence from exploration to validated concept, build, qualified release, launch, and post-launch learning. - Keep portfolio state legible: current stage, owner, next decision, dependency, evidence, budget assumption, and reason for continued investment. - Reduce bespoke founder coordination by turning repeated launch work into documented workflows that people and agents can execute. ### Turn ambiguous opportunities into bounded product bets - Clarify the user, problem, product promise, differentiated capability, adoption path, and economic hypothesis before allowing implementation to expand. - Distinguish facts, assumptions, forecasts, experiments, and unresolved decisions. - Convert broad initiatives into executable milestones with owners, dependencies, acceptance criteria, and stopping conditions. - Read architecture notes, API schemas, issues, pull requests, test output, and deployment evidence well enough to identify missing product decisions and unsupported completion claims. - Trace important user behavior across interfaces, services, data stores, integrations, and operational processes without needing to write the production code yourself. - Identify the smallest credible launch or market test that produces useful evidence. - Recommend continue, revise, defer, or stop based on results rather than sunk effort. ### Direct multi-project, agent-enabled execution - Create clear work packets for engineering, QA, research, design, growth, contractors, and AI agents. - Decide which investigations can run in parallel and which decisions must be resolved sequentially. - Maintain durable shared state so a product does not depend on one person’s private chat history. - Review agent-generated research, specifications, tickets, test plans, and launch artifacts for correctness, completeness, and unsupported claims. - Establish escalation thresholds so agents and contributors continue autonomously within scope but stop for missing authority, consequential ambiguity, security risk, or external commitments. - Use receipts, tests, deployed behavior, user evidence, and metric changes as the basis for completion—not confident summaries. ### Coordinate engineering and QA toward qualified releases - Work with engineering to translate product intent into implementation boundaries without pretending to be the architect. - Ask technically precise questions about interfaces, state, permissions, failure modes, observability, environments, and integration contracts. - Build acceptance criteria that cover user behavior, integrations, failure recovery, operational readiness, privacy, and security—not just the happy path. - Partner with the QA lead to create a repeatable qualification process across products. - Maintain separation between experimental, staging, and production environments and between demonstrated behavior and planned behavior. - Run release-readiness reviews and return a clear go, capped, held, or stopped recommendation with evidence. ### Own launch learning and portfolio economics - Define the target user, channel, activation event, retention signal, and commercial hypothesis for each launch. - Coordinate launch execution with the Product & Operations Lead and the existing marketing team when a product enters active go-to-market. - Establish simple, honest dashboards for acquisition, activation, use, retention, revenue, cost, and forecast-versus-actual performance. - Track time, external spend, infrastructure cost, opportunity cost, and expected value well enough to support portfolio decisions. - Design prediction-market, token, incentive, or community mechanisms only as bounded product hypotheses subject to technical, security, economic, and legal review. ## Initial areas of focus - Produce the first usable version of the launchpad operating model and apply it to a real product rather than writing a theoretical playbook. - Establish the intake, stage-gate, QA, evidence, and retrospective templates for subsequent launches. - Convert one broad portfolio opportunity into a product thesis, technical dependency map, smallest credible launch, and explicit decision gates. - Map which launchpad components can be reused across products and which must remain product-, chain-, or jurisdiction-specific. - Define the clean handoff between product qualification and active go-to-market ownership. ## What success looks like - Learn the active portfolio, technical systems, current team, and agent infrastructure. - Produce a concise portfolio view showing what is active, what is waiting, what is blocked, what should stop, and what decision is next. - Define and test a lightweight launchpad workflow on one real product. - Establish a dependable working relationship with your assigned agent and at least one specialist agent or execution cell. - Return one evidence-backed product recommendation that materially improves sequencing or prevents waste. - Move one product through the launchpad to a qualified launch or a well-supported stop/defer decision. - Make release criteria, QA evidence, launch dependencies, and post-launch metrics visible and reusable. - Demonstrate that a second product can reuse the operating model with less founder intervention and less reinvention. - Improve forecast accuracy around scope, time, launch readiness, and commercial outcomes. - Reduce portfolio thrash: fewer ownerless initiatives, fewer hidden blockers, and fewer projects that remain active without a clear next proof. ## The person we are looking for You are likely a strong fit if: - You have taken one or more zero-to-one products from ambiguity through launch, qualification, or a disciplined stop decision. - You can manage a portfolio without treating every opportunity as equally active or equally valuable. - You are technically fluent enough to reason with engineers about APIs, data models, authentication, state, AI systems, wallets, smart contracts where relevant, infrastructure, environments, observability, and integration risk. - You can inspect technical work directly and distinguish a product decision, an architecture decision, an implementation detail, a known limitation, and an unverified claim. - You can create strong product and release criteria without turning the role into ticket administration. - You understand experimentation, basic unit economics, product analytics, and the difference between a forecast and a commitment. - You can coordinate engineering, QA, growth, contractors, domain experts, and founders while keeping ownership explicit. - You are comfortable operating in crypto or other high-uncertainty markets without substituting hype for evidence. - You use AI agents as a serious operating system and can explain how you manage their context, tools, memory, permissions, concurrency, verification, and failure recovery. - You communicate clearly enough that someone can understand the state of a product, the evidence, and the next decision without attending every meeting. Experience with venture studios, developer platforms, productized services, marketplaces, prediction systems, Web3 infrastructure, or multi-product launch operations is especially relevant. Direct experience in every technical domain is not required, but you must demonstrate that you can learn a new system quickly and work credibly with its engineers. ## This role is probably not for you if - Your main PM strength is maintaining Jira, running standups, or relaying messages between specialists. - You treat a large backlog as evidence of a product strategy. - You cannot say no, defer work, or recommend stopping a weak initiative. - You need every requirement to be settled before you can design a bounded learning step. - You treat token issuance, market activity, or community attention as proof of product value by itself. - You use AI primarily to summarize meetings or draft tickets and have no deeper delegation or verification practice. - Your agent workflow is private, irreproducible, and impossible for another operator to audit or resume. - You believe agent-generated confidence substitutes for deployed behavior, tests, customer evidence, or professional legal/security review. ## Scope and decision rights You will have broad authority to structure product discovery, sequence approved work, define launchpad workflows, propose portfolio decisions, coordinate qualification, and recommend launch, cap, defer, or stop dispositions. You may delegate work within approved scope to your assigned agent and specialist agents. Delegating to an agent does not expand your authority. The founder retains final authority over company strategy, capital allocation, product selection, material spending, production releases, token or chain commitments, public financial claims, custody or execution of funds, legal and regulatory positions, and other irreversible or high-reputation decisions. Appropriate technical, security, economic, and legal reviewers retain authority within their professional gates. ## How to apply Send us: 1. A short note explaining why an AI-native, multi-product launchpad role fits your background and working style. 2. Two examples of products you moved from ambiguity to a launch, validated milestone, or disciplined stop. Include the original uncertainty, your decisions, the team, the evidence, and the outcome. 3. A sample artifact that shows how you think: a product brief, launch plan, dependency map, decision record, release checklist, experiment design, portfolio review, or postmortem. 4. A description of the AI agents you currently use. Name the products, models, frameworks, custom agents, automations, or orchestration tools where possible. 5. Walk us through one real multi-step agent workflow you operate. Explain the trigger, task decomposition, context sources, tools, memory/state, parallel work, approval gates, verification, and final handoff. 6. Describe one agent failure that mattered: incorrect research, lost state, premature completion, bad tool use, duplicated work, or an unsafe action proposal. How did you detect it, recover, and change the system? 7. Explain how you decide whether an agent-generated product recommendation is strong enough to influence engineering time or capital. 8. Tell us what you deliberately keep human-owned and what would have to become true before you delegated more authority. 9. Your location, working hours, availability, and preferred engagement model. We care more about evidence of judgment, launch ownership, technical fluency, and a real agentic operating practice than pedigree or polished job titles.