Founding Applied AI Engineer

Stackfix; Greater London; On-site; GBP 90,000 - 120,000; Full time

Benefits offered by this job

MacBook Pro

Equity

Private office in Holborn

AI model subscriptions

Job summary

Stackfix is hiring to own AI builds end-to-end for client projects. You will sit with operators on-site, map real-world workflows, and ship working software that integrates with existing client systems.

You will own the system after it goes live, focusing on reliability, data quality and measurable value. This is an on-site role in Holborn with equity and a MacBook Pro as part of the package.

Qualifications

• 4+ years building production software.
• Strong in Python.
• On-site role with operators.
• Ability to reason about reliability, evals, data quality and edge cases.
• Ability to create measurable value with software.

Responsibilities

• Own AI builds end-to-end for client projects.
• Sit with operators on-site and map how work actually happens.
• Build and ship working software.
• Integrate with the client’s existing stack.
• Own the system after it goes live.

Skills

Python

User interaction

Reliability & data quality

On-site collaboration

Tools

Django

HTMX

Alpine.js

TypeScript

Job description

We build production AI systems inside real businesses.

Examples: A factory processes orders by hand. A distributor loses deals because quotes take days. An international law firm has people copying data between systems.

We turn this manual work into production software that improves throughput, accuracy, and margin.

The role

You will own AI builds end-to-end for clients.

You will:

• Sit with operators on-site and map how work actually happens
• Build and ship working software
• Integrate with whatever the client already runs
• Own the system after it goes live

About us

Paddy (https://www.linkedin.com/in/paddystobbs) worked at Google, and later founded Jukedeck, a neural-networks-for-music company, which TikTok acquired.

Tom (https://www.linkedin.com/in/tomdekan) founded and sold an HR software company in 2023, and has 9+ years building AI systems, including using GPT2 when it was released.

We are backed by Seedcamp, Chalfen Ventures, and angels from DeepMind.

We use boring software wherever possible: our aim is to minimize complexity. The stack varies per customer. Our default stack is:

• Python
• Django
• HTMX / Alpine.js / Typescript

Who this role is for

We care about systems you have built, not job titles. You should apply if:

• You have 4+ years building production software
• You are strong in Python
• You like talking to users
• You can reason clearly about reliability, evals, data quality, and edge cases
• You care whether software creates measurable value

We also look for evidence of sustained excellence. The domain matters less than the pattern: high standards, long attention span, and visible output. Examples: Paddy (CEO) was previously a professional singer. Tom (CTO) played waterpolo for GB.

Who this role is not for

Please don't apply if:

• You want a remote-first role
• You don't want to understand how a client makes money
• You dislike talking to customers
• You want a narrow backend role or frontend role

This is for someone who wants work to be one of the main places they grow.

We optimise for autonomy, quality, and sustained output, not long hours.

What you get

• Ownership. You will own client builds from discovery to production.
• Direct commercial exposure. You will see how companies make money, where they leak margin, and where software creates value.
• Founder-level learning. You will work directly with Paddy and Tom, two exited founders, and see how we sell, scope, negotiate, build, hire, and run the company.
• Meaningful upside. £90,000-£120,000/year salary + equity + a new MacBook Pro + subscriptions to any AI models you need + a central private office in Holborn.

Interview Process

• Intro chat, 15 minutes - Fit, background, and what you are looking for.
• Technical screen, 60mins - A programming task (<60mins) and discuss it.
• Technical interview. - A realistic engineering discussion. No trick questions or leet code questions.
• In-person working session, 2-3 hours, in Holborn - A realistic problem based on the kind of work we do. No whiteboard theatre. No unpaid client delivery.