9Point Capital
Website:
9point.capital
Job details:
We're building a small, senior, AI-native team that ships vibe-coded apps, marketing content, and internal tools at a pace a traditional team can't match. We're looking for our first and most important hire — the person who sets the technical and AI bar for everyone who follows.
What this role is
You own what gets built and verify what ships. You define intent, set the architecture, and hold the review discipline that keeps an AI-first team from drifting into unmaintainable code. You're the human in the loop who turns fast AI generation into production-quality output. You'll likely have a voice in every hire after you.
This is not a "manage from a distance" role. You still read, critique, and architect code — you've just learned to orchestrate AI agents to handle generation while you concentrate your judgment on the decisions that actually need it.
What you'll own
- The studio's operating thesis: what we build, what we deliberately don't, and where the line sits between fast experimentation and regulated/production work
- Translating business goals into clear specs and architecture the team builds against
- The spec → plan → build → verify loop, and the final sign-off on every meaningful output
- The AI tooling and agent stack, chosen, standardised, and continuously upgraded as the landscape shifts
- Team culture, pace, and the hard calls on scope and trade-offs
The AI skill set we're looking for
- Agentic coding tools, daily and fluent : Cursor, Claude Code, Windsurf, or equivalent. You've shipped real things with these, not just demoed them
- Prompt and context engineering : you know how to feed an agent architectural context, reference existing patterns, decompose complex tasks, and constrain output to prevent dependency bloat and drift
- Agent orchestration : designing multi-step and multi-agent workflows where AI plans, writes, tests, and iterates under your oversight (not one-shot prompting)
- Model judgment : a working sense of which model fits which task: reasoning depth vs. speed vs. cost vs. context window, and when to switch
- Spec-driven development : writing specs and tests precise enough that AI-generated work stays verifiable and reviewable
- AI output review : the instinct to spot plausible-but-wrong code, hallucinated APIs, insecure defaults, and architectural drift before they compound
- Working knowledge of the API layer : calling model APIs directly, tool/function calling, structured outputs, and building AI features into the products themselves (not just using AI to build them)
- A point of view on where this is going : you track the tooling landscape because it changes monthly, and you upgrade the team's stack accordingly
Plus the engineering foundation underneath it
- Strong full-stack capability : you can build and debug front to back without AI when you need to
- A track record of taking products from idea to shipped, more than once
- Product taste, and the judgment to know when AI output is wrong or risky
- Comfort with ambiguity and a bias toward shipping small and iterating
Nice to have
- Background in cybersecurity, fintech, or regulated software
- Experience running a small, high-output team or studio
- A public portfolio of things you've personally shipped, bonus if any of it is AI-built
How we'll know it's you
In conversation, you can walk us through exactly how you'd ship a given product using AI agents the spec, the tooling, the model choices, the review gates, and the failure modes. You talk about review discipline and architectural drift without being prompted. And you have strong, current opinions about which AI tools to use and why.
If you architect, orchestrate, and review rather than write line-by-line, and you want to prove what a tiny AI-first team can do, let's talk.
Click on Apply to know more.