This is a sole-ownership role, not a task-execution role.
Why this role exists
Our product is a live, multi-tenant B2B SaaS that turns research and meeting evidence into sales and marketing assets: a Product Playbook, a Discovery Call Guide, a Deal Room, and a Content Plan. It runs on a patent-pending evidence-anchoring architecture so that every claim it generates can be traced back to a verbatim source. The product is in live alpha with real paying-track users.
The app has been built and operated through an AI-assisted development workflow. We now need a single, accountable human engineer to take full ownership: make it genuinely production-ready on an urgent timeline, put proper engineering and support practices in place, and own the architecture, cost, and roadmap from here. You will be the technical owner of the product. If something breaks at 9pm, it is yours. If the architecture needs to change to make the roadmap affordable, that call is yours to make and defend.
This suits someone who wants to own a real product end to end, not slot into a large team. You will have unusual autonomy and unusual responsibility.
What You Will Own
Production-readiness and QA.
- Stand up a proper staging environment. Today every change ships to production with no staging buffer. Fixing this is one of your first jobs.
- Own continuous integration and deployment. There is a CI pipeline (tests, linting, custom guards) to extend, and a publish process to make safe and repeatable.
- Build out the automated test suite (unit, integration, and end-to-end) and a manual QA process for what cannot be automated. Establish a regression suite so fixes stay fixed.
- Stand up observability and error tracking (error monitoring is scaffolded but not yet active), define what "healthy" looks like, and act on it.
Architecture and cost.
- Own the cloud architecture and keep it cost-effective while enabling the roadmap. The current stack spans a managed backend platform, an AI provider, a payments provider, and a low-code build-and-deploy tool. You will assess each, decide what to keep, and decide where to bring the stack fully in-house under standard infrastructure-as-code and CI/CD.
- Own and optimise AI spend. The product makes heavy use of large-language-model calls; model selection, prompt-caching, batching, and pipeline design materially affect the monthly bill. This is a real and ongoing cost lever.
- Decompose and harden the core generation service, which is currently a single large edge function, into a maintainable, testable architecture.
Security and data privacy.
- Maintain strict multi-tenant data isolation. The product relies on row-level security across every table; you must be fluent in this and able to verify it at runtime, not just on paper.
- Own compliance with the Protection of Personal Information Act (POPIA) and, for international users, GDPR. The product already implements user deletion, data export, consent records, and audit tombstones; you will maintain, formalise, and extend these, and keep us defensible.
- Run regular security reviews, manage secrets properly, and own the response if something goes wrong.
UI and UX.
- Clean up and raise the standard of the interface. Recent audits surfaced consistency debt (an inconsistent type scale, non-semantic colour usage, responsive gaps) alongside good bones. You will drive a design-system pass and keep the experience coherent as the product grows.
- Bring product sense, not just execution. You should be able to look at a screen and know when it is confusing a user.
User support and support workflow.
- Provide hands-on user support during alpha, and design the support function that outlasts the alpha: intake, triage, response targets, a feedback loop into the roadmap, and the lightweight on-call discipline a small SaaS needs.
Roadmap and process.
- Lay out, cost, and maintain the product roadmap with the CEO, and turn it into shipped software.
- Replace the AI-orchestrated development model with proper human-owned engineering practices: branching, review, release management, documentation, and runbooks.
The Stack You Will Inherit
- Frontend: React, TypeScript, Vite, Tailwind, shadcn/ui (Radix), TanStack Query, React Router. Vitest for tests, ESLint, GitHub Actions CI.
- Backend: Supabase (PostgreSQL with row-level security, Auth, Storage, and Deno/TypeScript Edge Functions). Scheduled jobs via Postgres cron for generation and email queues. Secrets in Vault.
- AI: Anthropic Claude, invoked through a multi-stage, tool-use pipeline with content-addressed evidence anchoring (patent pending).
- Payments: Stripe (webhooks with idempotency).
- Current build and deploy: a low-code platform (Lovable) currently manages the Supabase project and publishes the frontend; source lives in GitHub. Deciding the long-term future of this layer is part of the role.
You do not need prior experience with every item, but you must be able to own all of it within your first months.
You Are a Strong Fit if You Have
- 5+ years of professional full-stack engineering, with real seniority: you have owned systems, not just tickets.
- Deep TypeScript, and strong modern React.
- Strong PostgreSQL, including row-level security or equivalent multi-tenant data-isolation experience. This is non-negotiable; the product's security model depends on it.
- Production experience with Supabase, or with a comparable serverless-plus-Postgres or Firebase-style stack you operated yourself.
- Real cloud architecture and cost-management experience (any of Supabase, AWS, GCP, Vercel, Cloudflare). You can reason about what a design will cost at scale, not just whether it works.
- Practical data-privacy experience: POPIA, GDPR, or equivalent, applied in a real product (not just awareness).
- A track record of taking a product to and through production: CI/CD, testing discipline, observability, incident response.
- The temperament to be the sole or lead engineer: decisive, self-directed, comfortable owning the whole thing, and able to talk to users directly.
- Based in Cape Town, or able to work core hours in South African Standard Time.
Bonus points for:
- Building or operating LLM-powered products: Anthropic or OpenAI APIs, tool use, retrieval, prompt engineering, and evaluations.
- Experience replacing a low-code or AI-builder platform with a hand-owned stack without downtime.
- Stripe billing integrations.
- Deno and edge-function experience.
- A genuine eye for design and UX.
- Awareness of software patents and how implementation choices interact with filedclaims.
Why Join Us?
At Blacfox, we value our people and we value our business. We create a space where individuals can be authentic, grow in confidence, and build strong working relationships. We work hard, but we also make room for enjoyment—because when people feel safe, supported, and inspired, they do their best work.
How to Apply
Send a CV and a short note covering: a production system you have personally owned (what it did, your role, what you were accountable for when it broke), your experience with PostgreSQL row-level security or multi-tenant isolation, and one example of a cost or architecture decision you made and why. Links to shipped work welcome.
Work Location: Hybrid remote in Century City, Western Cape
