AI Agent Creator Fintech; aka Technical Product Manager

AI Agent Creator for Fintech (aka Technical Product Manager)

We’re building an AI-native banking platform. Built by AI agents and humans who keep up.

Location
: AI-first, Manus-first, Async-first, remote

The Role

You’ve built products for banks or fintech as an engineer. Became a PM. Now you can do both.

This isn’t product management. You’ll vibe code — ship banking prototypes in hours, make architectural calls real-time, write specs AI agents execute. AI increased the speed of making, not clarity of thinking. You bring both.

What you’ll actually do:

• Define roadmap with focus on banking/financial integrations
• Write specs that humans and AI agents execute flawlessly
• Design banking integrations, payment flows, compliance features
• Partner with legal, compliance, ops on regulatory requirements
• Communicate through memos and docs, not meetings

Language Mastery:
The Core Competency

In the AI age, proficient writers possess the most potential. Not about grammar—it's translating your vision into precise language that commands AI to build what you envision.

What we look for:

• Describe complex concepts, UX, or technical requirements in extreme detail (500+ words per feature)
• A pattern of reading, playing with language, or world-building (writing, D&D, creative projects)
• Write 10,000–20,000 word docs that are clear, structured, compelling
• Obsession with the perfect word or phrase to capture nuance
• Iterate on descriptions until AI outputs match your vision

Why this matters:

We build products where humans write specs and AI executes. Quality is limited by language clarity. What took 100 hours now takes 5 minutes—if the spec is precise.

If you “overthink” word choices or rewrite sentences to capture nuance, you might be who we’re looking for.

Our Spec-First, AI-Native Workflow

We’ve built a pipeline that changes what it means to be a PM:

The Core Loop:

• Automated Building
  : AI agents generate working code, designs, docs
• Video Review
  :
  Watch a video to verify it matches your vision
• Approval & Deployment
  :
  Approve, and the system creates PRs and deploys

Why this matters:

• Documentation IS the product. Everything flows from specs.
• In 3–4 years, AI-generated code will be too complex to read. The spec becomes source of truth.
• Not theoretical. We build this way today.

What we need from you:

• Write specs so clear AI executes without ambiguity
• Bias toward writing over talking
• Comfort that your primary output is documentation, not code
• Discipline to iterate until specs are perfect—precision saves exponential time

What “Vibe Coding” Actually Means Here

“Vibe coding” isn't writing sloppy code. It's a fundamental shift in how products are built:

The Old Way:

• PM writes a brief spec
• Engineer writes tech spec
• 100+ hours later, feature ships

Our Way:

• Write a detailed spec (with AI assistance)
• AI agents (Cloud Flow) read spec, find tools, generate code
• Review a video of what was built
• Approve and it creates a commit

What this means for you:

• Spend 80% writing and refining specs, not coding
• Language quality directly determines output quality
• Be good enough at coding to review AI code and spot issues
• Be exceptional at describing UX, edge cases, and requirements in writing
• Iterate on descriptions, not code

Example from our workflow:

A team member wrote 400+ words describing a logo—“soft L‑shaped modules,” “eyes vertically centered,” “no eyebrows, no mouth,” “medium orange rounded rectangles.” AI generated exactly what was envisioned. This precision is what we mean by “vibe coding.”

Who You Are

Required:

• 5+ years as engineer in banking/fintech (non-negotiable—deep domain expertise)
• 3+ years in product, business, ops, or non‑engineering roles
• Understand code well enough to build prototypes and review AI code
• Hands‑on experience with banking products
• Founder experience a plus
  :
  Built a fintech product from scratch, even as side project? Strong signal

Technical:

• Experience with microservices and distributed systems
• Constantly searching for new ways to build with AI
• Build websites with AI generators (Lovable, Replit, Claude)
• Build small internal apps

Product:

• Write specs that serve as source of truth
• Shipped dev‑facing products or APIs
• Data‑driven with qualitative judgment

The “Vibe Coding” Mindset:

• Prototype…