Independent technology advice for organisations with too much manual work, too many systems and no appetite for expensive guesswork.
Alchementary is an independent UK technology practice helping organisations turn operational problems into proportionate, useful solutions.
Sometimes the answer is AI. Sometimes it is an integration, a form, a process change or ordinary software. Sometimes the best advice is not to spend the money.
The commercial front door
Before you buy a platform, commission a build or announce an AI initiative, understand where the value actually is.
We examine how work moves through your organisation, identify the opportunities worth pursuing, estimate their likely value and difficulty, and give you a practical order in which to tackle them.
See what is includedDifferent starting points, one shared principle: understand the real need before reaching for the technology.
Solution design
Once you know what is worth doing, we can describe what to build or buy: requirements, architecture, security boundaries, costs and a route to implementation.
Our products
We create and operate focused digital products of our own—born from real problems, shaped by real use and improved over time.
Prototype sprint
We prototype a fixed question against a fixed dataset, with a clear definition of success and failure. A useful result can be “do not build this”.
Independent judgement
We untangle technical problems, pressure-test plans and work alongside your team when you need experienced, independent thinking.
AI onboarding
We help businesses and individuals find meaningful uses for AI, adopt it responsibly and build the habits needed to keep getting better.
Each one started as a problem we could not buy a good answer to. All four are documented, in use, and honest about their limits.
4ensics Leviathan
A complete binary analysis, reverse-engineering and digital-forensics workbench, running inside an ordinary browser tab — with your artefacts kept on your own machine.
Blackwall
An OAuth 2.1 and OpenID Connect authorisation server where every human signs in with a passkey — and every action leaves a trail that can prove it was not altered.
Fury
Post-quantum file encryption, a documented file format and a Go library, for material that must stay unreadable long after the mathematics has moved on.
NeoCMS
Controlled visual editing for the static website you already have — with drafts, revisions, scheduled publishing, media and SEO tools, while public pages remain ordinary HTML.
We work in short, legible stages. You will know what we are learning, what we are making and why it matters.
Understand the people, the pressure and the real job to be done.
Turn ambiguity into a sharp brief, clear choices and an honest measure of success.
Prototype early, test with reality and build only what earns its way forward.
Leave you with a useful result—and the confidence to keep making it better.
“Use ChatGPT” is not an AI strategy.
The value comes from knowing where AI fits, where it does not, and how to combine it with your expertise. We help you move beyond isolated experiments to useful, repeatable practice.
Map the work, identify strong use cases and prioritise opportunities by value, feasibility and risk.
Learn prompting, verification, data handling and human oversight in the context of work you actually do.
Explore reusable workflows, custom assistants, knowledge tools, integrations and automation where they add value.
Create sensible guardrails, internal champions and measures that turn early wins into lasting capability.
Code an AI helped write still has to survive somebody hostile. We run an adversarial, assume-breach audit that maps the system the way an attacker would, then works through authentication and authorisation, injection and browser security, cryptography and key management, secrets across their whole lifecycle including in memory, native memory safety, denial of service and resource exhaustion, dependencies and supply chain, infrastructure, frontend robustness, business-logic abuse, code-quality risk and the gaps in your security tests.
Every finding is grounded in the actual code or configuration, and carries severity, confidence, the evidence itself, a realistic attack scenario, impact, remediation, suggested tests and the relevant framework mappings — no speculative hardening and no scanner noise. You get a security audit report and a separate reliability, maintainability, testing and UX report, and then we fix what is confirmed: the smallest safe change, regression and abuse-case tests where they earn their place, the checks actually run, and only passing fixes committed. Your existing work is preserved, and anything that cannot be fixed safely is written down rather than quietly left.
Available for
IndividualsTeamsOrganisations01You know the problem, but not what the answer looks like.
02The available tools nearly fit—but “nearly” is costing you.
03You need an honest technical view without a sales agenda.
04Your AI experiments are interesting, but not yet transformative.
05You value thoughtful work, plain speaking and practical outcomes.
Tell us about the problem, the idea or the change you are trying to make. It does not need to be a polished brief.
[email protected]