Employer: Gaia Learning
Location: UK-based, remote-first. Cheshire/North West preferred
Working hours: full-time
Salary: £60,000
-
The role at a glance
Location: UK-based, remote-first. Cheshire/North West preferred
Type: Permanent, full-time
Salary: £60,000
Holiday: 31 days including bank holidays
Reports to: Chief Technology Officer
Checks: Right to work in the UK, criminal record self-declaration, online screening, and post-offer pre-employment checks (including an Official Enhanced DBS with Barred List Check). See “Why UK-only, and the background check” below.
About us
Gaia Learning runs Bloom, a live online education platform for young people with SEND and EHCPs. It’s used by teachers, learners and school administrators, with guardian access coming.
Lessons are taught live in the browser: video, a shared whiteboard, resources, attendance and an end-of-lesson review, backed by learner profiles, objectives, safeguarding records and AI-assisted summarisation of documents and lesson outcomes.
We’re past MVP and into production, working with schools and local authorities, iterating in sprints. The platform is small, modern and unusually well documented for its age. It now needs a senior engineer alongside the CTO to own it properly.
The role in one line
You’ll be the senior engineer on a production platform, writing most of the code across the frontend, the API and the cloud infrastructure that runs them, and owning the quality bar for a product holding children’s special category data.
This is a hands-on, high-ownership role, not a management one. You’d take primary day-to-day ownership of the codebase and pipeline, and set the standards the rest of the work is held to.
If you want to be the person who knows the whole system end to end and is trusted to make architectural calls, this is that job. If you want a narrow ticket queue inside a large team, it isn’t.
What you’ll own
The product code. Full-stack feature work across a React and TypeScript single-page app and a PHP API. Most features touch both sides, so you’d design the API and build the interface that consumes it.
The infrastructure. Our cloud environments and the infrastructure-as-code that provisions them, with the security and data-residency controls a platform holding children’s data requires.
The pipeline. Test-gated builds and deployments, and the alerting around them.
The quality bar. Testing, type safety, accessibility and security standards across the codebase. You’d set them, raise them where they need raising, and hold the rest of the work to them.
The stack
React and TypeScript on the front end, PHP on the back, MySQL, all running on Google Cloud in a UK region and managed as infrastructure-as-code. There’s real-time collaboration and AI-assisted summarisation in the product.
You won’t be expert in all of it on day one, nobody is. We care that you can become expert in the parts you aren’t, and we’ll go through the architecture in detail at the technical stage.
How we build: agentic coding
We build with Claude Code, and this role requires you to be genuinely on board with that. Our repository is set up for it, with guardrails, written architecture rules and an explicit definition of done, and a meaningful share of our output comes through agentic workflows. It’s not a side experiment; it’s how a team this size ships at the pace we do.
We’re not looking for someone who’ll tolerate it. We’re looking for someone with their own workflow, real opinions about where these tools help and where they don’t, and the discipline to review AI-written code as rigorously as a colleague’s, more so, because the failure modes are easier to miss.
We’re looking for someone who has
Essential
Substantial experience building and running production web applications, with real depth on both sides: not a frontend engineer who has seen an API, or a backend engineer who can force a React component into existence
Strong TypeScript and React: component architecture, state, forms, performance, and the discipline to keep a feature-based codebase from decaying into a shared-folder swamp
Strong server-side experience in PHP
Relational data modelling: design a schema, write a safe migration against live data, reason about a query under load
Cloud infrastructure ownership and infrastructure-as-code. Comfortable being called when production is unhappy, and comfortable saying the fix is in Terraform rather than clicking around a console
Security as a habit rather than a phase: every endpoint authenticated and authorised, input validated server-side, secrets out of the client, least privilege by default
Fluent, considered use of agentic coding tools, Claude Code specifically, or clearly transferable experience
Clear written communication. You’ll work with a small non-engineering team, external suppliers, and schools
Desirable
WebRTC or real-time collaboration
Accessibility remediation or formal audit experience
UK GDPR in practice: special category data, DPIAs, data residency, retention
LLM integration beyond a demo: evaluation, injection defence, cost and latency, and a clear head about what these systems shouldn’t be trusted with
EdTech, SEND, safeguarding, or another regulated domain
Not required
A degree, a particular set of prior employers, or a specific number of years. Show us you can build the thing well.
How we work
Small team, low ceremony, high standards. Sprints, a ClickUp backlog, PRs reviewed before merge, and a written Definition of Done: typecheck and build clean, backend tests green, style clean, authorisation covered, manual keyboard and accessibility pass on any changed UI.
Architecture rules are written down and enforced. You’re welcome to argue with any of them, in writing, and change them. Documentation is part of the change, not a follow-up ticket.
How we work together
Remote-first, UK-based. We’d like you within reach of Brighton, where you’d co-work with the CTO one or two days a week. Early-stage architecture benefits enormously from a whiteboard and a coffee, and this is the stage where that pays off most. If you’re commutable rather than local, let’s talk.
Compliance, security and accessibility
Bloom is built for schools and local authorities, so we’re assessed before we’re bought. Compliance isn’t a separate function here. It lands in the codebase and the infrastructure, and it would be part of this role.
We handle special category data about children under UK GDPR. Data residency, encryption, access control and auditability are enforced by policy rather than convention, and DPIAs, retention and lawful basis are live engineering constraints rather than paperwork. Our security work is shaped by ISO 27001 and Cyber Essentials.
Accessibility matters here more than in most places, as our learners are the reason WCAG exists. An accessibility pass is part of our definition of done, and holding the platform to WCAG 2.2 AA is part of this role.
None of this is box-ticking. It’s the reason schools let us near their learners.
Why UK-only, and the background check
Bloom holds special category data about children with SEND: EHCPs, health information, safeguarding records. Everything is UK-hosted and pinned there by policy, and production access is tightly scoped.
This role involves production access, so it carries strict UK residency and right to work in the UK.
Shortlisting and pre-interview screening: Shortlisted candidates will be asked to complete a criminal record self-declaration and undergo online background and social media screening in line with KCSIE statutory guidance prior to interview.
Conditional post-offer vetting: Formal offers of employment are subject to the satisfactory completion of post-offer pre-employment checks, including: inspection of an Official Enhanced DBS with Barred List Check (original paper certificate); overseas criminal record checks or certificates of good conduct for candidates who have lived or worked outside the UK for 6 months or more (continuously or in total) within the last 10 years; medical fitness verification for the post; and verification of essential qualifications and two satisfactory professional references.
If any of these fundamental eligibility or safeguarding checks are a blocker for you, we’d rather know now than at offer stage.
It also sets the honest framing for the job: the consequence of sloppy work here isn’t a bad sprint. We want someone who finds that motivating rather than intimidating.
Process
Intro call: 30 minutes, with the CTO.
Interview 1, technical conversation: 60 to 90 minutes. A walk through the real architecture, the decisions we’ve actually made, and what you’d have done differently. A look at some code of your own choosing, or a system you’ve built if your work is all proprietary.
Interview 2, non-tech and cultural conversation: 30 to 60 minutes. Meet with the CTO and CEO to discuss life at Gaia Learning, Bloom and the wider team.
Final conversation: terms, offer.
We’ll come back to everyone who applies.
The Work From Home Hub is dedicated to finding and sharing remote and hybrid job opportunities across the UK. The content available on theworkfromhomehub.co.uk is intended for general informational purposes only. While we strive to ensure the information provided is accurate and reliable, we make no guarantees or warranties, expressed or implied, regarding the accuracy, adequacy, validity, reliability, availability, or completeness of the information presented on our site.
Whether you are looking to change careers, advance in your current role, navigate workplace challenges, or return to work after a break, I can give you expert guidance and practical strategies to help you achieve your goals.
Visit Victoria-Hopkins.com for more information
This page may contain adverts for products and services tailored to those seeking homebased roles or already working remotely. These adverts help The Work From Home Hub source and share more job opportunities to our community. Support our mission - become a partner today.