AI-native SEND research and software company

£10 billion spent annually on SEND. Zero evidence of what actually works.

We are building the research, the platform, and the agentic AI to change that. Safe by design. Evidence by default. Built for 638,745 children who deserve better than paperwork.

638,745

Children with active EHCPs in England. Growing 10.8% every year with no sign of slowing.

46.4%

The proportion meeting the 20-week statutory deadline. The lowest compliance rate ever recorded.

99%

How often councils lose at SEND tribunal. Not contested cases. Systemic failure on paper.

This is not a funding problem. It is an evidence problem.

The UK spends more on SEND than almost any comparable system. The issue is that nobody knows whether the money translates into better lives for children.

£10.54B

Spent with no feedback loop

Annual high-needs spending across England. Not a single structured dataset connects this money to whether children's lives improve.

DfE High Needs Block 2024/25
53.6%

Plans that arrive too late

Over half of new EHCPs miss the statutory deadline. At the worst councils, the figure drops to 1%. Children wait months or years for support that should take 20 weeks.

DfE SEN2 2025 Release
£153M

Burned on fights councils keep losing

22,276 tribunal cases in 2024. A 43% surge. Councils lost 99% of those that reached hearing. These are not close calls. They are plans that never should have been issued.

HMCTS Tribunal Statistics 2024; IPSEA
~£6B

Conditional on proof of improvement

The government is absorbing council SEND deficits. The condition: submit reform plans that demonstrate how outcomes for children will get better. Not efficiency. Outcomes.

DfE February 2026 Announcement

Research, software, and agentic AI designed around one question

Does the support in this child's plan actually improve their life? Everything we build exists to answer that.

1

Professional reports flow in

EP assessments, medical reports, parental statements. Our AI extracts and structures the clinical findings automatically.

2

Personalised drafts, not generic outputs

RAG architecture retrieves relevant case law and guidance at generation time. No child data touches the model. Every statement is sourced and cited.

3

The golden thread, validated

Graph-based algorithms check that needs connect to outcomes connect to provision. Vague language gets flagged before it becomes a tribunal case.

4

Outcomes tracked across years

Annual reviews feed structured data back into the system. Over time, the platform answers which provisions produce results and which do not.

We are not building a faster pen. We are building the feedback loop that never existed.

The SEND industry has spent a decade digitising paperwork. Plans get drafted faster. Deadlines get tracked. But no one has built the infrastructure to measure whether the support prescribed in those plans actually changes a child's trajectory.

Sensol is an AI-native SEND company. We research provision effectiveness, build software that structures SEND data for the first time, and deploy agentic AI that coordinates the multi-agency process safely. The savings come not from cutting corners, but from finally knowing where £10 billion is well spent and where it is wasted.

RAG Architecture Golden Thread Validation Outcomes Tracking Parent Portal Multi-Agency Orchestration Children's Code Compliant

Three things no existing SEND tool does

Competitors validated that councils pay for EHCP automation. Good. We build what they left out.

📊

The outcomes dataset that does not exist yet

No platform in the UK links specific SEND provisions to whether children actually improve. We are building that dataset. The White Paper now requires councils to prove outcomes to access £6 billion in funding. This is not a feature. It is infrastructure they need by law.

👨‍👩‍👧

Families see what is happening for the first time

Every existing tool faces the council only. Parents are locked out. Our portal gives families real-time visibility into plan progress, plain-English explanations, and provision status. The adversarial dynamic that drives £153 million in tribunal costs starts with information asymmetry. We end it.

🔒

AI safety built into the engineering, not the policy page

We process children's disability data. That demands more than a privacy notice. Our middleware enforces the ICO Children's Code at the technical layer: purpose-coded data tagging, algorithmic transparency records for every AI output, and automated screening that prevents the AI from disadvantaging the child it serves.

A lawyer, a SEND caseworker, and an engineer at Goldman Sachs walk into a problem

This is not a punchline. It is the team composition this market actually requires.

YN

Yvonne Wambui Nyangi

Founder & CEO

LL.M International Economic Law (Queen Mary). Legal advisory at The Commonwealth. Compliance at Clifford Chance. Frontline carer for individuals with learning disabilities and complex needs. FDI Moot champion (All Africa).

CU

Caylem Urssher

Co-Founder & COO

SEND domain specialist with frontline experience across three local authority SEND teams: Lambeth, Newham, and Sandwell. Assessment caseworker, annual review tracking, EHCP quality assurance. Knows every workflow from the inside.

DO

Daniel Wole Olamide

CTO

Software Engineer, Core Engineering (Authentication) at Goldman Sachs. Built RAG-powered AI systems. Led replacement of 20-year-old authentication system across 800+ applications. BSc Informatics & Computer Science, Strathmore University.

What we are thinking about, investigating, and building

Analysis of SEND policy, AI in public services, and the evidence gaps we exist to fill.

📋
SEND PolicyComing Soon

The White Paper rewrote the SEND market in one afternoon

Every Child Achieving and Thriving introduces ISPs, retains EHCPs, and ties £6 billion in deficit relief to outcomes evidence. What this means for councils, schools, and companies like ours.

Read article →
🔬
ResearchComing Soon

638,745 plans and no one asked if they worked

Castro-Kemp et al. found most EHCP outcomes are not functional. Ahad's review found no systematic tracking of provision effectiveness. The gap is not accidental. It is structural.

Read article →
⚖️
AI & EthicsComing Soon

What "safe AI" actually means when you process children's disability data

Everyone claims privacy by design. We explain what that looks like at the engineering layer: purpose-coded tagging, transparency records, and detrimental use screening that runs on every output.

Read article →

We are looking for pilot councils, co-design partners, and investors who think the SEND system deserves better evidence

Whether you work in a SEND team, run a school, care for a child with SEND, or invest in companies that solve hard problems.

Sensol is pre-launch. We are in the research and build phase. If you have experience, opinions, or questions about what we are doing, we want to hear them.

🏢
CompanySensol Limited · England & Wales
🌐
4