Ai Automation & Full-Stack Product Engineer

 

Description:

CerealBox is looking for its first in-house AI Automation & Full-Stack Product Engineer: someone who can take a real business problem, figure out the smartest technical solution, build it, launch it, maintain it and continuously improve it as better AI models and tools become available.

This isn’t a role where someone gives you a list of tickets every morning.

You’ll be expected to ask:

“Why are we doing this manually?”

“Can AI solve this?”

“Should we build it ourselves or integrate an existing tool?”

“How do we make this 10x better?”

And then actually make it happen. 🚀

 

A little about CerealBox

We’re a UK-based creative and performance agency working exclusively with charities.

We manage digital fundraising and paid media for around 30 charity clients and have already built our own analytics platform, True ROAS, to connect advertising spend with actual donation revenue.

Now we want to take the next step.

We want AI and automation embedded across the business not because it sounds cool, but because it can genuinely help our team work smarter, move faster and deliver better results for charities.

 

Your first big responsibility

You’ll take ownership of True ROAS, our existing production platform used by our media buyers, creative team and leadership.

Today, it already handles:

• Attribution between Meta Ads and donation data
• ROAS reporting by client, campaign and creative
• Thousands of creatives and performance insights
• Commercial and settlement reporting
• Scheduled data syncs and API integrations
• Currency conversion and system monitoring

You’ll maintain it, improve it and help decide what it becomes next.

 

Then comes the fun part 🤖

Some of the problems we want you to explore include:

• AI Copywriting: Can AI consistently produce strong advertising copy?
• Creative Generation: Can static ad concepts and designs be automated?
• Script Writing: Can AI develop better hooks, concepts and video scripts?
• UGC: How much of UGC production can be generated using AI?
• Video Editing: Can AI take raw footage and produce high-quality finished ads?
• Media Buying: Can an AI agent help launch, monitor and optimise campaigns?
• Creative Ideation: Can AI generate genuinely interesting campaign concepts rather than generic ideas?
• Landing Pages: Can we rapidly generate and optimise fundraising landing pages?
• Reporting: Can reporting become almost completely automated?
• Internal AI Assistant: Can we make tools like Toast smarter so our team can simply ask questions and get useful answers?
• Client Communication: Can AI safely respond to common or repetitive queries?
• Workflow Automation: Can we remove repetitive operational work across the company?
• Meta + LLM Integrations: How can Meta data interact with models such as Claude and other LLMs?

You won't necessarily build every idea exactly as described.

Part of your job is deciding what is actually worth building.

 

What you’ll do

 

🧠 AI & Automation

• Identify repetitive work that can be automated
• Research AI tools, APIs, models and agents
• Prototype solutions quickly
• Decide whether we should build, integrate or buy
• Turn successful experiments into reliable production tools
• Keep our AI systems current as models and technologies evolve

 

💻 Full-Stack Engineering

• Build and maintain modern React and TypeScript applications
• Own backend architecture, APIs and database logic
• Work extensively with PostgreSQL and Supabase
• Build reliable third-party integrations
• Maintain scheduled jobs and data pipelines
• Monitor failures and make sure systems are reliable and secure