Written by the Head of Development at MeritDocs
I lead development at MeritDocs, so I was not reading these discussions for clever keyword ideas. I was looking for the point where a real school workflow starts to break. Reddit did not write our product brief, but it gave us real examples of where special educational needs and disabilities (SEND) paperwork becomes too large, too repetitive, or too difficult to keep current. We turned those examples into test cases and asked whether we could reduce repeated admin without taking professional judgement away from school staff.[1][2][3]
When you are building software, it is easy to mistake a clean demo for a solved problem. A generated paragraph can look fine in isolation. The harder question is whether a special educational needs coordinators (SENCOs) can trace it back to evidence, edit it, find it later, and trust the version they are looking at. Public forum discussions are not a national workload survey or a customer testimonial, but they often describe that friction in much plainer language than a software requirements document.
There is an important legal boundary behind this work. In England, the local authority secures preparation and maintains the statutory education, health and care (EHC) plan. Schools contribute evidence, advice, pupil and family views, provision information and review material. MeritDocs supports that school-side evidence and document workflow; it does not issue statutory Education, Health and Care Plans (EHCPs).[8]
The first test cases came from messy real-world problems
One public discussion described a sudden workload spike and a large bundle of documents due before parents' evening. It gave us a useful workload pattern.
Another commenter in the same discussion describes an earlier paperwork system as long and slow, then says a shorter template in the school's management information system (MIS) made the work quicker, although it still took a large amount of planning, preparation and assessment (PPA) time.[1]
That became one of our first product questions. If a SENCO already has the relevant evidence, can our system help them produce a focused, structured document without turning every plan into another long form to maintain?
The lesson was simple: SEND software needs to make busy periods easier to manage without losing sight of which record staff are reviewing.
A different discussion, Overwhelmed with SEND, describes annual EHCP reviews eating into PPA, repeated short-turnaround forms, and the cumulative pressure of many small requests.[2] One comment also points to a common source of duplication: the same school-wide adjustment being recreated by multiple classroom teachers instead of being made available once by the SEND team.[2]
This is where the implementation gets less glamorous but more important. Useful information should be reusable, but reuse cannot mean copying old wording blindly into every new document. We needed the system to carry relevant evidence forward while leaving staff to decide what still applies.
Annual reviews showed us where the workflow becomes fragile
The clearest example came from a discussion about leading EHCP annual reviews. A primary teacher describes being asked to monitor review dates, prepare documents, arrange meetings, contact agencies, complete the annual review, and prepare the school's evidence and proposed updates for the EHCP with pupil voice, progress, and new targets.[3]
The post describes more than ten hours of additional work in one setting when several EHCPs were involved. Another commenter describes an annual review as potentially taking a full day and an EHCP application as taking ten or more hours when evidence gathering is included.[3] Those are individual accounts, not national statistics. They still expose the shape of the problem.
An annual review is not just a form to complete. It involves dates, people, evidence, progress, family views, pupil voice, decisions, and what happens next. A useful tool has to keep those parts connected. It should make the current information easier to assemble and review, not imply that software can make the professional decision for the SENCO.
For EHC plans, the statutory responsibility remains with the local authority. Section 37 of the Children and Families Act 2014 requires the local authority to secure preparation of the plan and maintain it once prepared. Schools contribute evidence, advice and review information; MeritDocs supports that school-side workflow rather than issuing the plan.
The Department for Education's workload advisory group made a similar point at system level, focusing on the unnecessary workload associated with data management and evidence collection.[4] The SEND Code of Practice also makes clear that support and review sit within a wider process involving children, young people, parents, and relevant professionals.[5]
We tested the record, not just the generated paragraph
From the developer side, the most important lesson was that a fluent paragraph is not the same as a useful SEND document. The output has to survive the rest of the workflow. Someone needs to check it, understand where it came from, find it again at the next review, and know whether it is still current.
We tested the workflow against questions such as:
- Can staff start with evidence they already hold instead of a blank page?
- Can information be reused without losing its source or context?
- Can a SENCO see which document is current and when it needs review?
- Can an authorised member of staff find the right pupil information quickly?
- Can pupil, parent, teacher, and professional contributions remain visible?
- What happens when an old Portable Document Format files (PDFs) or Word document does not fit the school's current structure?
- Does the system show uncertainty, or does it quietly invent a tidy answer?
A separate discussion about school information systems asked what teachers would improve in a combined platform. One commenter described several steps needed to find a quick summary of a class's needs and open pupil passport documents.[6] We treated that as a usability prompt, not as a reason to criticise a named product. If a member of staff is authorised to see a pupil's current support information, the interface should not make them remember a maze of clicks to get there.
What we built into the product
MeritDocs is a UK SEND documentation and records platform. We built it around the artificial intelligence (AI)-enhanced SEND part of the workflow because generating text was never the whole problem.
The Document Wizard helps staff create structured SEN Support Plans and profound and multiple learning difficulties (PMLD) plans from recorded pupil information and relevant UK SEND material. Staff review the result before it becomes part of the record, because the system cannot decide whether a pupil's needs have changed or whether a target is right.
The Documents Hub is the other half of the product. It keeps SEND documents in one searchable place, with filtering, visible review dates, and export. We kept coming back to a simple test: could an authorised member of staff find the current record without asking one person to remember where it lives?
We also had to deal with the unglamorous reality of old PDFs, Word files, and photographs of paper plans. The import process is deliberately best-fit rather than magical. UK SEND documents vary between local authorities and schools, so imported content may need checking. We would rather show what mapped, what needs review, and what did not fit than present a polished document and quietly lose evidence.
You can read more about the wider workflow in How SENCO software should reduce paperwork without losing accuracy and Why unmapped content matters when schools import SEND documents.
What we deliberately did not promise
The discussions also stopped us from making some easy but stupid promises. Software cannot fix a staffing model, make statutory decisions, or turn a difficult pupil situation into a tidy dashboard.
The product does not replace SENCO judgement. It does not guarantee statutory compliance, perfect imports, or a particular inspection outcome. It does not turn a staffing problem into a software problem that disappears after a demo.
Our promise is narrower. We want a school to be able to create, find, and review its records with less dependence on one person's memory. The professional work still belongs to the people who know the pupil, the family, and the context.
Office for Standards in Education, Children's Services and Skills (Ofsted)'s Supporting SEND report found variation in how needs and provision were understood and delivered. It also reported that some SENCOs did not have enough time to carry out their responsibilities and faced bureaucracy around referrals and EHC plan assessments.[7] Reddit is not proof of the scale of that problem, but the practical language in these discussions helps explain what that pressure can feel like on the ground.
The test any SEND software should pass
Whether a school uses this platform or another system, the useful questions are straightforward:
- Can staff see the current document without searching through multiple folders and inboxes?
- Can the system reuse evidence while showing what still needs professional review?
- Can it preserve pupil and parent voice rather than flattening everything into generic language?
- Can it handle local variation and show unmapped content during migration?
- Can the school export its records and leave the system without being trapped by it?
Those questions are more valuable than an impressive AI demonstration. A SEND tool should make the record more dependable and the work more usable. It should not create another dashboard that staff have to maintain for the sake of the dashboard.
That is what the public discussions helped us test. Not whether AI could write a convincing paragraph, but whether the whole workflow could be made clearer, more reviewable, and less fragile for the people responsible for supporting pupils every day. That is a much harder engineering problem, but it is the one worth solving.
Sources
[1] TeachingUK discussion on individualised education plan (IEP) and provision-map workload [2] TeachingUK discussion on SEND workload [3] TeachingUK discussion on EHCP annual reviews [4] DfE teacher workload advisory group report [5] SEND Code of Practice [6] TeachingUK discussion on school information systems [7] Ofsted Supporting SEND report [8] Children and Families Act 2014, section 37