# uk / software / it
UK consultancy. Experienced developers for internal tools, infrastructure, large deployments, AI-assisted builds, and product launches. Security is part of the job. Pricing is scoped. Support continues after go-live if you want it.
$ man engagement
People who have shipped software and then had to keep it running. Engagement size follows the risk: a contained build, a longer platform job, or a launch that needs someone around afterwards.
Internal tools, automations, customer-facing web apps, the kind of software staff use every day. We keep the scope written down and pick a stack your team can maintain without calling us for every change.
Examples: admin panels, reporting jobs, APIs between vendors, first versions of a product, rebuilds of tools that outgrew a spreadsheet or a brittle prototype.
Cloud layout, multi-service platforms, reliability work, and integrations between systems that were never designed to meet. We advise when you want a second pair of eyes. We build when you need the work done.
Common ground: containers, serverless, VMs, Kubernetes, Cloudflare Workers and other edge runtimes, databases, queues, CI/CD, logging and metrics. We plan for growth instead of assuming traffic stays flat.
Apps and portals from brief to live. We stay involved in product choices that change delivery cost or risk, so launch week is not a scramble through unfinished tickets.
Before go-live we sort environments, secrets, rollback, monitoring, and who to call in the first weeks. A green pipeline alone is not enough.
$ ls capabilities/
Broad deployment experience is the default. If a platform is outside what we know well, we say so before anyone starts cutting over production.
Most of the usual deployment shapes: bare metal and VMs, Docker and Compose, Kubernetes, AWS and similar managed cloud, Cloudflare Workers/Pages, hybrid setups, and promotion between environments.
Threat modelling, code review with security in mind, auth and access control, dependency and secrets hygiene, hardening, and fixes you can actually ship. We treat this as normal engineering work, not a deck of slides.
Multi-tenant systems, busy APIs, databases under load, caching, queues, and the day-to-day ops habits that keep alerts useful when traffic jumps.
Model APIs, retrieval over your own data, evaluation, and basic guardrails. We use AI to speed engineering up when it helps, and to build product features when the problem suits it. We avoid unverifiable magic in production.
After launch: incidents, small improvements, dependency updates, and a route to someone who already knows the system.
$ cat billing.md && cat support.md
Quotes against a written brief. Fixed price when the scope is clear. Time-and-materials when we are still discovering the work. No invented multipliers and no multi-year contract required to get a normal rate. You know the commercial shape before we start.
We can stay on after launch for systems we built, or for systems we have taken over carefully: watch the dull checks, fix what breaks, plan upgrades without forcing a rewrite.
Sensible defaults, reviewed auth, tidy dependencies, clean deploys. If you need a formal audit and that is outside the engagement, we say so. Either way we leave the system safer than we found it where we touched it.
Used in engineering and in product work when it is the right tool. We do not paste secrets into random chat tools. Model calls in production get logging, limits, and a way to turn them off.
$ cat /etc/company
No matches.