Whater.org helps organisations embed a reflective local layer that protects sensitive legal and business data on-device, satisfies data-residency and consent requirements by design, and still uses your best models on the market.
Or follow the mission. No tracking pixels, no marketing funnel.
Layers between AI and us. Privacy at the edge. Governance on demand.
Privacy is architecture, not a policy document.
Imagine if AI can be edited, someone decides its morals, ethics and behaviour. You are the architect of your life, your health, your morals, and your AI's behaviour. LocalLayer is how you build it.
To serve humanity by creating AI related tools and services that enhance people's agency. The current default for AI is “send everything to a server and trust us.” That was never good enough for us, and it should not be good enough for your AI questions, your medical records, your family's information, or anything you or your organisation has a duty to protect. We intend to build in value: your organisation's agency, and your clients' personal agency.
Those five words hold everything else together. We would rather be honest about what is unfinished than dress it up as more than it is. Collaborate with us.
We work with corporates today, building LocalLayer to the same exacting standards of transparency that agencies like CMS and the FDA are asking of for instance, health AI SaaS, because the people behind the data deserve nothing less. However, we go further. Sensitive records stay local. Top-tier AI models still work on the data. Governance is built in, not bolted on, including a versioned consent gate before health or biometric data reaches any model.
LocalLayer is not just software sitting between you and a model. It is the place where a human stays in the loop, where the AI checks before it answers, and where every decision leaves a trail you can audit. Set out in our first paper, EveR Local Layer (v2.3.1, September 2025), with more of the reasoning still being written up.
We ship tools that prove the architecture in public.
RedactUS strips personal data before any prompt leaves the device. Open-source and free.
AI Doctor Ben is the first paid implementation of the governance layer, and also Whater.org's first partner in the work of understanding health signals responsibly.
(RedactUS Pro, a version for the legal profession is in private development.)
Concierge-level health clarity for you and your family. Search gives you scary answers. Your context with us changes everything. Make sense of your family's lab results, privately on your device.
Upload lab results. Multiple AI perspectives, conventional, integrative, functional, and more, from one place.
A developing timeline of your ECG, cholesterol, hormones, supplements, all on your device.
Built by Whater.org. Powered by LocalLayer.
AI Doctor Ben is an exploratory health insights platform and does not provide medical advice, diagnosis, or treatment.
A citizen science initiative into the health signals of tomorrow. His signals. Her signals. Our future.
A grand mission, stated plainly: build the world's most useful, most privately governed picture of non-medical health signals, and put it to work for public research, not private profit alone. Growth comes first, contribution follows once the platform earns your trust. More soon.
Whater.org partners with corporates, governments, and grant programmes to deploy a version of LocalLayer inside their environment. The pattern is the same in every engagement: sensitive data stays local, top-tier models still work, governance is built in, and the person at the centre, citizen, employee, patient, child, keeps control.
Healthcare, legal, finance, HR, any organisation that wants AI working on records it cannot send to the public cloud. We deploy and configure LocalLayer in your environment, document the controls for your regulator, surface Article 9 consent gates where health or biometric data is involved, and train your team to operate it.
Governments everywhere are being asked to prove that AI serves citizens without exposing them. Whater.org gives you a working answer to both questions at once: engagement without surveillance, and a paper trail your regulator can actually read.
Health insurers, corporate wellness divisions, and assurance providers can offer AI-backed guidance to members and employees without ever holding their raw health data centrally. LocalLayer keeps the sensitive record on the person's device, not in your warehouse.
partner@whater.org goes to a real human, usually within one working day.
We are early conversations with university teams on the science behind His & Hers and the wider Whater.org research agenda. We are actively looking for academic partners in the United States and the United Kingdom who want to shape how citizen-contributed health signal data is collected, governed, and used for public research.
If you work in public health, biostatistics, data governance, or a university business development office and this sounds like your world, we would like to hear from you.
The current default for AI is “send everything to a server and trust us.” You know that problem of another server breach, another hacked server. All are caused by your data going to a cloud service. Through direct development, that was not good enough for us, and it is not good enough for your AI questions, personal data, medical records, family information, child-safety contexts, or anything an organisation has a client duty or regulatory duty to protect. So we are building the alternative: a local redacting, reflective layer. Our services and tools are proven in public, and we run a partner programme for the organisations that need it at scale.
An AI platform is different, expansive, globally available, accessible to the right sort of people. LocalLayer is built to give rise to tools other people can build on: engineers adding their own mark, companies setting their own standard of ethics, legal teams using our redaction service to run their own conversations with any LLM they choose. That is how you measure impact, not by what one team builds, but by how other people go on to transform with it.
For that reason, we expect people, companies, and governments to make this their own: to edit the behaviour, ethics and morals layer that governs how AI treats them, whilst we manage the AI side.
The work is led from the UK by a team with an e-Health UK national award and a Qualcomm Tricorder XPRIZE quarter-final result behind it. Delivery is supported by the mentorship of an AI-listed company head, and for our health SaaS, a shadow council of retired medical consultants who can speak freely of the health service they wish existed with AI. The partner page is the right door for corporates, governments, and grant programmes. Also all Medical Consultants, working or retired, in the US or UK, may also contact us to be considered for our councils. The newsletter is the front door for everyone else.
Paper 1. The EveR Local Layer: A Reflective, Local Self-Training Layer for AI Assistants (v2.3.1, September 2025). The engineering, published. Read →
The governance and public-facing arguments are still being written. We would rather publish them once than publish them early and revise in public.
RedactUS, open-source, on GitHub, live today. See the repo →
Further tools are in internal development and will be released when they are ready to stand on their own.
AI Doctor Ben, the first paid implementation of LocalLayer, live today.