About Returning.

AIThink Discord, spin-the-wheel, lucky draws, streaks, rewards, and leaderboards - but built for financial institutions.

Returning.

AI builds gamified loyalty and retention systems for CFD brokerages. Our clients are enterprise brokers paying $60-144K per year. We are a 10-15 person team based in Singapore, profitable, and growing fast.

Our product lives inside broker client portals and helps brokers drive trader engagement through rewards, challenges, lifecycle campaigns, referrals, trading incentives, and gamified experiences.

This is not generic SaaS.

We work with real integrations, messy client systems, compliance-sensitive environments, production constraints, and commercial stakes. The product is niche, fast-moving, and execution-heavy.

Why this role existsWe are looking for a product manager who can use AI to help build product faster.

Not someone who just writes tickets.

Not someone who only runs meetings.

Not someone who talks about AI strategy from a distance.

We need someone who can take messy product inputs, use AI tools aggressively, create structure, prototype flows, pressure-test ideas, inspect technical constraints, and help move work from raw concept to shipped product.

The next version of our product team will not be built around heavier process. It will be built around better orchestration: humans, engineers, designers, QA, docs, AI agents, prototypes, specs, and feedback loops all moving together.

Your job is to help run that operating loop.

What you will actually doUse AI as your product operating systemYou should be deeply comfortable using tools like Codex, Claude Code, Cursor, ChatGPT, or similar to think, research, prototype, analyse, structure, test, and produce working artifacts.

This role is for someone who can use AI to multiply execution, not someone who occasionally asks ChatGPT to rewrite a paragraph.

You might use AI to:- turn a messy product idea into a clear product brief- inspect a codebase or API documentation- prototype a new client portal flow- compare implementation options- generate edge cases and QA checklists- draft technical handoffs for engineers- investigate whether an idea is feasible before engineering time is spent- create internal tools, scripts, mockups, or structured docs- compress hours of scattered context into clear next actionsAI should be part of how you work every day.

Turn messy inputs into buildable workYou will take founder ideas, client requests, sales feedback, broker constraints, Slack threads, call notes, rough sketches, and half-formed product thoughts and turn them into something the team can actually build.

That means clear specs, flows, priorities, acceptance criteria, edge cases, and tradeoffs.

You should be able to bring structure without creating bureaucracy.

Prototype and test ideas directlyYou do not need to be a full-time engineer, but you should be technically comfortable enough to make things real.

You should be able to use AI coding tools to mock flows, inspect frontend behavior, test APIs, wire lightweight internal tools, or create working prototypes that help the team understand what should be built.

The goal is not to replace engineering. The goal is to reduce ambiguity before engineering starts.

Bridge product, design, and engineeringYou will work closely with designers, engineers, QA, and commercial teams.

You should be able to explain business intent clearly, understand technical constraints, catch vague requirements early, and help the team find the smallest useful version that can ship.

When a requirement is unclear, duplicated, overbuilt, or commercially weak, you should catch it before it becomes rework.

Understand technical systems well enough to be usefulOur product depends on broker portals, APIs, authentication, event payloads, trading data, rewards, campaigns, and integrations.

You should be comfortable reading API docs, understanding JSON, tracing request flows, inspecting payloads, testing endpoints, and asking engineers specific questions.

You do not need to be a senior developer.

You do need to be technically dangerous enough to reduce confusion instead of adding to it.

Own follow-throughImportant product decisions should not disappear into calls, Slack threads, or someone's memory.

You should make sure decisions become clear actions, open questions are resolved, blockers are surfaced, and shipped work matches the intent.

Must haves- Singaporean or PR- Degree in Computer Science, Business, Entrepreneurship, or a related field- Strong hands-on usage of AI tools such as Codex, Claude Code, Cursor, ChatGPT, or similar- Strong written and verbal communication- Comfortable reading technical documentation, JSON, API flows, and system diagrams- Able to turn ambiguous product problems into clear next steps- High agency and strong follow-throughWhat good looks like- Product ideas move from messy input to buildable scope much faster- Designers and engineers receive clearer briefs with fewer missing pieces- AI meaningfully improves product velocity, not just productivity theatre- More ideas are prototyped before engineering time is committed- Product tradeoffs surface earlier- Fewer avoidable misunderstandings become rework- The team ships faster without becoming process-heavyWho we are looking forAI-native in practiceYou already use AI tools as a serious part of your workflow.

You can show how AI changes how you research, write, prototype, inspect systems, make decisions, and produce work. You are not just "interested in AI." You have changed how you operate because of it.

Product-mindedYou can decide what the smallest useful version is. You know when to push for polish and when to ship. You can connect product decisions back to business value, client reality, and execution cost.

Technically comfortableYou can read docs, understand API behavior, reason about data shape, test things yourself, and hold useful conversations with engineers.

You do not hide at the "stakeholder management" layer when the details matter.

Strong communicatorYou write clearly. You summarise well. You can turn a confusing discussion into a useful product brief. You can explain technical constraints without making them sound more complicated than they are.

Good under ambiguityYou are comfortable walking into a half-formed situation and creating structure.

You do not need every task to be perfectly scoped before you can contribute.

Creative but groundedYou can spot non-obvious solutions, combine ideas, and use AI to explore possibilities quickly.

But you are also practical about tradeoffs, timelines, engineering capacity, and what the team can actually maintain.

What this role is not- Not a traditional PM role built around ceremonies and status reporting- Not a project coordinator role whose main job is chasing updates- Not an AI strategy role- Not a role focused on building LLM features for customers- Not a pure engineering role- Not the right fit for someone who needs neat requirements and clear swim lanes before startingStrong signals- You have used AI tools to produce real work, not just content- You have turned vague business ideas into shipped product or working prototypes- You have worked in a startup or founder-led environment- You can read technical docs and narrow ambiguity yourself- You can show examples of specs, prototypes, workflows, or tools you created- You are comfortable moving between product, technical, and commercial detailsWeak signals- You mostly talk about frameworks, rituals, and stakeholder management- You need polished tickets before you can be useful- You avoid technical detail- You say you are "into AI" but cannot show how it changes your workflow- You create process as a substitute for judgment- You want to manage product from a distance instead of getting close to the workWhat you get- A high-ownership role in a profitable, fast-growing startup- Direct exposure to real enterprise clients and production constraints- A team that already uses AI heavily and expects you to do the same- A role across product, design, engineering, integrations, and execution- The chance to help shape how an AI-native product team actually operatesHow to applyIf you made it this far, send an email to desmond[@]returning.ai and you will get a confirmed response.

Include:1. Why product management is the path you want2. What AI tools you use and how deeply you use them3. One example of something you used AI to build, prototype, analyse, or ship4. One example of a messy situation you turned into a clear outcome5. How you would rate your people skills, honestly6. Your MBTI, if you have done oneWe will respond to every serious application. No ghosting.