Clients and prospects have been asking whether Servent Notes could support the EHCP process. Rather than give them a generic “yes”, we looked at where the product could genuinely help, where it shouldn’t, and why annual reviews look like the strongest place to start.
718,838
Active EHC plans in England in January 2026
597,000
Plans expected to have an annual review during 2025
339,000
Annual reviews recorded as completed during 2025
200,000
Completed annual reviews resulting in a decision to amend
Over the past few months, we’ve been asked variations of the same question:
Could Servent Notes support EHCPs?
It’s an understandable connection to make.
Servent Notes captures conversations and turns them into structured, reviewable records. Education, Health and Care Plan processes depend on information from children and young people, families, schools, health, social care and local authority teams.
But EHCPs are statutory documents. The processes around them involve professional judgement, multiple organisations and sensitive information about children and young people.
So “yes, AI can do that” isn’t a particularly useful answer.
We looked instead at the EHCP process and asked a different question:
Where does the existing capability of Servent Notes fit naturally, without asking AI to do a job that belongs to a SEND professional?
Our current view is that EHCP annual reviews are the strongest place to start.
Why annual reviews?
There were 718,838 active EHC plans in England in January 2026, up 12.5% in a year.
Around 597,000 plans were expected to have an annual review during 2025. The Department for Education recorded 339,000 as completed. Of those completed reviews, 200,000 resulted in a decision to amend the existing plan.
Source: Department for Education, Education, health and care plans, reporting year 2026
That is a substantial recurring workflow.
An annual review isn’t simply a form to complete. It considers whether the needs, provision and outcomes in the existing plan remain appropriate. A formal review meeting forms part of the statutory process, and the outcome must ultimately be communicated to the parent or young person within 4 weeks of that meeting.
That meeting can bring together the child or young person, parents or carers, education staff and representatives from health, social care and the local authority.
People discuss what has happened since the previous review. Progress against outcomes. Changes in needs. Whether existing support is working. New aspirations. Concerns. Further evidence that may be required. Actions and recommendations.
Then the meeting ends.
Somebody still has to turn what happened into an accurate record.
That’s the part of the process that interests us
There are already specialist products using AI elsewhere in the EHCP lifecycle.
Agilisys, for example, has an established EHCP drafting product and is now validating an Annual Review Agent with councils. Its annual review product pulls together existing progress data and professional input and produces structured review documentation for SEND officers to review.
Invision360 also has specific technology supporting EHCP annual-review processes and quality assurance.
Those products helped clarify the question for us.
We don’t need Servent Notes to become another EHCP drafting system.
The more natural fit is earlier:
Capture what happened in the conversation, structure it while the detail is fresh, and give the professional a useful record to review.
Think about what happens after the meeting
A review may have involved several participants and covered several sections of an existing plan.
The person responsible for the documentation needs to establish what was said about each outcome, what the child or young person communicated, what parents or carers raised, what professionals advised, what has changed and which actions were agreed.
Local authority processes illustrate how quickly that information then needs to move. Councils commonly require the review report and associated documentation to be circulated and submitted within 2 weeks of the meeting, before the local authority reaches and communicates its decision within the statutory 4-week period.
Source: Kent County Council, EHC plan annual reviews
So there is a fairly concrete administrative job between:
“The meeting has finished”
and:
“We have an accurate review record ready for the next stage.”
That is where we think conversation intelligence could help.
What could an AI-assisted annual review look like?
With the appropriate recording, lawful-basis and information-governance arrangements agreed, the annual review conversation could be captured as it happens.
Speech could be transcribed with individual speakers identified. The conversation could then be analysed against the authority’s agreed annual review structure.
Instead of beginning the write-up with a blank document, the practitioner could receive a structured draft.
What happens during the review | What the draft could structure |
|---|---|
Existing outcomes are discussed | Progress against each outcome |
The child or young person contributes | Views, wishes and aspirations |
Parents or carers contribute | Parent/carer views |
Needs have changed | Evidence of potential changes |
Existing support is considered | Evidence about whether provision remains appropriate |
New support is discussed | Proposed changes for professional consideration |
Information is missing | Further evidence required |
Actions are agreed | Action, owner and target date |
The important word is draft.
The professional reviews it, corrects it and decides what belongs in the formal record.
The benefit isn’t that AI makes the decision.
It is that the practitioner starts with a structured representation of the meeting rather than having to reconstruct it afterwards.
Could it also identify where the existing plan may need attention?
This is another area we think is worth testing.
The Department for Education recorded a decision to amend the plan following 59% of annual reviews completed during 2025.
Rather than attempting to rewrite the EHCP automatically, the output from an annual review could include a working change schedule.
Existing plan area | Evidence captured at review | Follow-up |
|---|---|---|
Special educational needs | A change in need was discussed | Consider whether Section B requires amendment |
Outcomes | Progress or a change in priority was recorded | Review the existing outcome |
Provision | Current support was discussed | Assess whether provision remains appropriate |
Child/young person’s views | New views or aspirations were captured | Update the relevant record |
Professional evidence | Further input is required | Obtain named advice |
Actions | A next step was agreed | Record owner and target date |
The schedule would tell the case officer where to look.
It would also preserve the distinction between evidence captured during the review and the professional or statutory decision made afterwards.
That distinction matters
This is where we would draw a firm boundary around the use of AI.
The annual review meeting contributes evidence and recommendations. The local authority remains responsible for deciding whether an EHC plan should remain unchanged, be amended or cease to be maintained.
The professionals involved remain responsible for their professional advice.
The practitioner completing the record remains responsible for checking that it accurately represents what took place.
And the child’s or young person’s contribution needs particularly careful treatment. Technology can help capture and organise what they communicated. It shouldn’t reinterpret their views and present that interpretation as theirs.
A useful system therefore needs more than good generated text.
The person reviewing the record needs to be able to check where important information came from.
Why the technology architecture matters here
The conversations involved in an EHCP annual review can contain information about a child or young person’s education, disability, health, social care, family circumstances and safeguarding.
Recording the meeting creates an additional sensitive data asset.
That means a council needs to understand where audio is stored, where transcription takes place, which AI services process the content, who can access it, what is retained and how generated outputs can be checked.
This is one reason Servent Notes’ deployment model becomes particularly relevant.
Servent Notes is deployed as a dedicated instance within the customer’s own Microsoft Azure subscription. Recordings, transcripts and model calls remain within that environment. The platform uses Azure AI Speech and Azure OpenAI, and outputs are versioned rather than simply overwritten.
It can also provide citations from generated answers back to the relevant speaker and timestamp in the recording, allowing somebody reviewing the output to return to the source conversation.
Native mobile applications and encrypted offline recording are also available today, which may matter where professionals are capturing interactions away from an office or reliable connection.
Those are existing product capabilities.
An EHCP annual review workflow is not.
That distinction is important too.
Capture: Annual review conversation
Structure: Draft against agreed review format
Review: Professional checks and corrects
Continue: Approved evidence enters the existing SEND process
So, can Servent Notes support EHCPs?
Our answer today is:
We think Servent Notes could support part of the EHCP process well, with annual reviews looking like the strongest starting point.
We would not position it as a replacement for a SEND professional, a local authority’s statutory process or a specialist EHCP drafting platform.
The specific opportunity we want to test is much narrower:
Capture an annual review conversation once, turn it into a structured and traceable draft, and give the professional a better starting point for the documentation that follows.
The sensible way to establish whether that actually works isn’t a large technology programme.
It is one annual review workflow.
Use the existing process. Use the authority’s existing documentation. Work with a small group of appropriate practitioners. Agree the information-governance approach. Capture the meetings and compare the resulting drafts with the current way of working.
We would want to know how long it currently takes to move from a completed annual review meeting to a usable record, how much practitioner time is involved, what needs correcting in an AI-assisted draft, whether the required information is captured consistently and whether professionals trust the output enough for it to be useful.
If it works, there is a basis for doing more.
If it doesn’t, we learn where the limitations are before attempting to embed it deeper into a statutory process.
This started with a customer question
That’s worth returning to.
We didn’t begin by deciding that Servent Notes needed an “EHCP proposition”.
Clients and prospects asked whether the product could support EHCPs.
Looking properly at that question has led us to a more specific answer than the one we might have given initially.
The interesting opportunity isn’t necessarily writing the EHCP. It may be capturing and structuring some of the evidence the process depends on, at the point that evidence is generated.
Annual reviews give us a defined, repeated workflow in which to test that idea properly.
So we’ll end with the same question we’re now asking the SEND teams we speak to:
How long does it currently take your team to turn an annual review meeting into completed, usable review documentation?
If you are responsible for EHCP annual reviews and are willing to compare your current process with this approach, we’d like to hear from you.
More from the blog:
