|
• AI-ready content architecture: lessons broken into granular chunks that can be refined and edited by AI based on individual student outcomes, opening the way to an AI-driven content feedback loop • Lesson content separated from visual layout, opening the door to reuse across Braille, plain-text and future formats • Manual production-design step removed from the publishing workflow • First product (Breaking News for students) live in January; flagship curriculum product (ULS) on track for beta later this year • Architecture chosen for developer self-service so Everway's own technical team can extend the platform without ongoing 67 Bricks involvement • Lead engineer embedded post-build to protect momentum and transfer knowledge cleanly to incoming contractors |
ing contractors
| Client |
Everway (formed by the merger of N2Y and Texthelp) |
| Region | UK and US |
| Sector | EdTech: K-12 and accessible learning content |
| Ownership | Private equity-backed (Five Arrows) |
| Engagement | Discovery, platform evaluation, CMS build, embedded handover |
| Duration | 12 months, from discovery through to handover |
Everway is an educational technology company formed by the merger of N2Y, a US business, and Texthelp, a UK-based assistive-technology specialist. Together they improve outcomes for students with a range of educational needs, and support the teachers who work with them.
Much of their work centres on creating accessible learning content for the US curriculum, with a particular focus on students with special needs. The N2Y side of the business, which 67 Bricks worked most closely with, produces lesson materials and lesson plans designed to let students with special needs learn alongside their peers, in the same classrooms, taught by mainstream teachers.
"Their real focus is on education and supporting students, and the teachers who support them, across a range of educational needs. Creating accessible content that can be used to educate, upskill and support students." — Rachael Lammey, Engagement Lead, 67 Bricks
Lesson content sat at the heart of Everway's business, both teacher-facing materials explaining how to deliver a lesson and the student-facing assets shown in class. But the process for producing that content had calcified around tooling built around fifteen years earlier.
Authors wrote lessons in Google Docs. A separate production-design team then rebuilt each document by hand in a legacy visual CMS that worked by dragging and dropping boxes on top of one another to assemble a finished page. The process was slow and manual, and it fused content and layout together. The same lesson text could not easily be reused in another format. Plain text, Braille, or any variant stripped of the visual styling, all required rebuilding from scratch.
"When the systems don't work, people find ways to work outside the systems. That creates inefficiencies and wrinkles, and things that are harder to work with downstream as part of a publishing process." — Rachael Lammey, Engagement Lead, 67 Bricks
The deeper problem was a ceiling on what Everway could build next. Creating a new template imposed enough cost on the production-design team that authors reused existing templates simply to avoid the work. Version control was absent: the system held hundreds of near-identical lesson revisions that nobody could track.
Everway's content team, many of them former teachers with deep interest in pedagogical research, could see better ways to teach. The tools made it hard to act on them. The manual pipeline also blocked the AI-enabled workflows Everway wanted next: breaking lessons into granular content chunks that could be refined and edited by AI based on individual student outcomes, with feedback from educators and assessment engines flowing back into the platform. Without a modern content architecture, a full AI content feedback loop was out of reach.
Everway had attempted an equivalent project in-house roughly a year earlier and scrapped it. The team got bogged down trying to solve every problem up front, without a clear way to break the work into prioritised, deliverable pieces. They ran out of capacity to make decisions. A small internal development team, already committed to other work and operating in the middle of a two-company merger, had no room to take on another major build.
67 Bricks was already on Everway's radar. The two had spoken around a year earlier about a different piece of work and, although that pitch was not successful, the relationship stayed warm. When this project emerged, Everway returned to 67 Bricks. The right specialist capability arrived at the right moment for both sides.
67 Bricks began with an unusually thorough discovery phase. The hardest part of the project was not implementation but reconciliation: there were many competing requirements, all of them important, with no two teams agreeing on the priorities. Content authors who understood teaching wanted one thing; the design, web and publishing specialists wanted another. Early work focused on aligning those views into a single product and getting people across the organisation talking to one another.
This is where 67 Bricks acts as a critical friend, pushing back on assumptions rather than building whatever a client first asks for. Everway had arrived with a list of specific requests that were several steps ahead of the problem. The team challenged those assumptions and worked back to the real underlying needs before getting into the details of a solution.
With the real needs surfaced, 67 Bricks weighed three routes: build something bespoke, buy an off-the-shelf product, or buy a platform and customise it. The team evaluated candidate CMS platforms, spoke to suppliers and ran user testing with Everway's authors. The analysis was shared openly so Everway could own the final decision.
They landed on customisation. The transparent evidence-led process, rather than a default to 'we will build it for you', resulted in a platform with its own support and maintenance, no lock-in, and a faster route to value.
The chosen CMS framework was selected for flexibility, customisability and developer accessibility. It is driven by code schemas held in version control rather than locked behind a fixed UI. The build deliberately decoupled lesson text from its visual presentation, restoring the separation the legacy system lacked and opening up content reuse across formats and products. Critically, the new architecture breaks lessons into granular content chunks: the foundation for AI-enabled personalisation and refinement based on individual student outcomes.
Tim Barclay led the early proofs of concept and investigations. The 67 Bricks team of Tim Barclay, Stephen Nichols, Tim Crilly and Oksana Horlock built the platform between September and December, interfacing with Everway's infrastructure team and working alongside a project manager from the client.
Everway has a capable in-house technical team, which let 67 Bricks choose a platform that is self-service for developers: any competent engineer can read the schemas and extend the system. The choice trades a small amount of non-technical click-through convenience for substantially more long-term power. The platform remains fully customisable as new requirements emerge, with no ongoing dependency on 67 Bricks.
When the core engagement ended in December, the rest of the 67 Bricks team rolled off and Tim stayed on in a consultative lead-engineer role. He was embedded alongside a team of incoming contractors and Everway's own project and engineering managers, with one job: hand over cleanly without losing the thinking or the momentum the project had built. This continued-consultancy-into-handover model is one 67 Bricks has used with other clients and continues to favour.
The CMS was one part of a much larger technical overhaul. Other teams were rebuilding the application teachers and students interact with, the systems that report on student progress, and the mapping of skills to the many US state educational standards: an area governed by genuinely complex regulatory frameworks. The biggest risk was the scale of simultaneous change.
Compounding that, requirements were not settled at the outset. It was not that Everway did not know what it wanted. Different people held different views, with little consensus beyond the high-level gist. 67 Bricks treated this as an organisational risk and met it with a single engineering principle: prioritise flexibility. No decision could be assumed final, so whatever was built had to absorb change without breaking.
The groundwork for faster production is in place, along with something the old system never offered: the capacity to iterate on what Everway delivers as new needs emerge, and an AI-ready content architecture that supports personalisation and refinement of individual content chunks based on student outcomes.
The first product to go live, in early January, was Everway's Breaking News offering, which publishes simplified versions of news stories for school-age students. Their flagship curriculum product, ULS, is due to reach beta later this year, with a full release the following September. Much of the value will be realised as the platform supports large volumes of back-to-school content, moving the production process from half-manual to fully automated for that cohort, and unlocking the AI-driven content feedback loop that the legacy system made impossible.
One clear signal: Everway was able to retire the entire production-design function that previously rebuilt Google Docs into published pages by hand. Concrete performance metrics will follow as more of the platform comes into service.
"It has opened up the capacity to speed that work up, and the capacity to iterate and add new features in future, which their old system did not have." — Tim Barclay, Lead Engineer, 67 Bricks
A discovery-led start that drew out the real requirements before any production code was written, and aligned stakeholders who had been pulling in different directions.
A transparent build, buy or customise evaluation that put Everway in control of the decision and avoided lock-in to a single supplier.
An end-to-end 67 Bricks team covering product, project management and engineering, so recommendations were implemented by the people who made them.
An AI-ready content architecture designed to support personalisation and refinement of individual content chunks based on student outcomes.
Technology choices anchored on flexibility, control and agility, the qualities Everway needs to keep evolving as their wider overhaul continues.
A deliberate handover model with a lead engineer embedded to transfer knowledge and protect momentum once the core team rolled off.
| Will Bailey |
Originated and brokered the client relationship |
| Rachael Lammey | Engagement lead and discovery |
| Alex Howat | Product and project management; sprint and delivery coordination |
| Tim Barclay | Lead engineer; proofs of concept, build and embedded handover |
| Stephen Nichols | Software development |
| Tim Crilly | Software development |
| Oksana Horlock | Software development |