Senior Full-Stack Engineer; R&D & Core Platform

Founder & CEO at AUDITSU | Building the compliance layer for digital accessibility

Location:

Leeds (City Centre) – Hybrid (3 days onsite / 2 remote) Salary:
Up to £70,000 + Stock Options

We are AUDITSU. We are building the engine that automates accessibility testing for the enterprise.

We aren't looking for a ticket-mover. We are looking for a Foundational Engineer
.

This is a role for an engineer who is tired of moving buttons 3 pixels to the left. If you want to solve graph traversal problems, handle raw mobile device streams, and architect a system from “ambiguous experiment” to “stable product,” this is for you.

We have the infrastructure (Django, Redis, Appium, Next.js). We have the funding. We have 1100+ clients on our waitlist. Now, we are rebuilding our core proprietary technology: a traversal engine that can autonomously navigate, interact with, and map complex Android applications.

You will act as the bridge between R&D concepts and production reality. You won’t be buried in a squad; you will work directly with the Founder and CTO with high autonomy.

We operate on a clear partnership model:
Product-Led, Engineering-Solved. The Founder defines the “What” (Top-Down product concepts and user goals), and you own the “How” (Bottom-Up architectural strategy and execution). We don't hand you tickets; we hand you problems, and we trust you to think critically to engineer the solution.

We need someone who is “full stack” in the truest sense. You must be comfortable diving into low-level Appium automation scripts, designing the Postgres schema that maps the data, and shipping the React UI that visualizes it.

• Architect the State Engine
  :
  Mobile accessibility testing fails because it lacks context. Your first challenge is to build a system that turns raw, noisy device streams into a deterministic, queryable graph of the application.
• Solve the “View” Problem
  :
  Mobile viewports only show a slice of the application. You need to engineer a method to capture, normalize, and reconstruct a complete “map” of a screen’s state, overcoming the limitations of scrolling and dynamic loading.
• Build the Decision Logic
  :
  Prototype the intelligence that drives the bot. You will determine how the engine handles loop prevention, detects screen changes, and decides whether to tap, scroll, or back out.
• Production Integration
  :
  Ensure this new engine isn't a silo. It needs to speak fluently to our existing FastAPI/Django backend and stream results to the Next.js dashboard in real-time.

Current platform stack (what’s already live)

• Next.js 14, Type Script, Tailwind
• Digital Ocean today, likely moving towards GCP

Traversal engine

• Rebuilt from scratch. You’ll help choose the right approach and tools, then integrate it into the platform above.

• You are a First-Principles Thinker
  :
  You don't just glue libraries together. You understand why things work. Whether you write the code yourself or review what an LLM generated, you can spot the race condition in the database lock or the inefficiency in the HTTP request immediately.
• You treat AI as a Force Multiplier
  :
  We don't care if you use Copilot/Cursor to write the boilerplate. In fact, we expect you to leverage AI to move faster. But you aren't a “vibe-coder,” you have the deep expertise to audit the output, refactor the architecture, and own the final implementation.
• You ship complexity
  :
  You have experience taking a messy, complex problem (like video streaming, large-scale scraping, or automation) and turning it into a clean, stable API.
• You value “Done” over “Perfect”:
  You can prototype fast (with or without AI assistance) to validate a hypothesis, then ruthlessly refactor to harden it for production.
• You are language-agnostic but Python-native
  :
  You produce clean, typed, testable Python. You are comfortable jumping into Type Script/React to build the tools you need, focusing on the outcome rather than the syntax.

• Experience with Appium
  , Selenium, or UIAutomator
  2.
• Experience with Computer Vision (OpenCV, stitching logic).
• Ba…