Engineering the systems business runs on

TEKWROUGHT modernises, builds and operates business-critical software for UK organisations. Fixed fees from a measured survey.

Book an infrastructure diagnosisSee our work

Survey

We measure the estate before we quote. Out-of-support software, slow paths and security gaps are found and costed.

  1. 01 Survey
  2. 02 Rebuild
  3. 03 Run

TEKWROUGHT keeps business-critical software running. We survey what you run, rebuild it without downtime and operate it under a service level agreement, for organisations where an hour offline has a price.

We measure before we touch anything

The console is reading this website from your browser as you look at it: how long a request takes, where it travels and which security controls are in place on the response.

Every engagement starts with the same instrumentation pointed at your systems. It gives both sides a factual baseline before a fee is agreed, and it stays in place after go-live so performance is reported to you every month.

  • Response times and their spread, from outside and from your own logs
  • Runtime, framework and library versions against their support dates
  • Certificates, security headers and exposed services
tekwrought.co.uk · live
round trip, median
…
95th percentile
…
origin first byte
…
security headers
…
  1. Waiting to start. The console runs when it scrolls into view.
Reading page delivery

Where we work

Sectors where software failures stop operations. Reference architectures show how we would engineer the core system; delivered platforms are live and operated by us.

Reference architecture

Logistics and cold chain

Live temperature telemetry across a fleet, with excursions caught in seconds.

Reference architecture

Manufacturing

One scheduling and uptime view across the shop floor, in place of spreadsheets.

Reference architecture

Financial services

A credit engine that shows its working, fed by an encrypted document portal.

Restoring Lives Worship Center — build preview

Delivered

Charities, faith and education

Public site, management system and installable app on one codebase.

Academy of Influence International — build preview

Delivered

Higher education

Admissions, student portal, examinations and registry for a theological college.

What we do

We take on three kinds of work. Each is scoped from a survey of what you run today and priced as a fixed fee, so the number you approve is the number you pay.

Your core platform runs on software that is out of support. Every release is a risk and the people who understood it have moved on.

Core Systems Migration

Legacy estates moved off out-of-support PHP, jQuery and ageing server stacks onto a low-latency Next.js and PostgreSQL architecture, without a dark day for the business.

scope
Legacy estate → Next.js · PostgreSQL
billing
Fixed fee · 30 / 40 / 30 milestones
deliverables
Target architecture · migrated services · runbooks
acceptance
Record counts reconciled and signed off · performance budget met

Sales, operations and finance work from different systems and different numbers, and the gaps are filled with spreadsheets.

Bespoke Operations Core

A unified operating system for the business, running unattended in the cloud: customer front ends wired natively to the databases, ERPs and shop-floor systems behind them.

scope
One operating system across front end, ERP and MES
billing
Fixed fee per phase · milestone billing
deliverables
Data model · integrations · scheduled workflows · audit trail
acceptance
Each phase run against your own data before sign-off
See Genesis

The system works, but no one is accountable for keeping it fast, patched and available.

Enterprise Infrastructure SLA

Continuous hypercare for systems that cannot go down: monitored, patched, scaled and improved by the engineers who know them.

scope
Production estate under hypercare
billing
Monthly retainer · agreed engineering hours
deliverables
Patching · query tuning · capacity reviews · quarterly report
acceptance
Response windows and uptime targets as agreed in the SLA

Start with an infrastructure diagnosis

Two to three weeks with read-only access to your logs, monitoring and code. You receive a costed account of what is wrong, the target architecture and a phased, fixed-price roadmap your finance director can approve. From £3,500 depending on the size of the estate, and credited in full against the build if you proceed within 60 days.

Book a diagnosisOr run the free website check

How we deliver

Every engagement follows the same four stages, from a two-week diagnosis to a multi-year platform programme.

  1. Survey

    A factual baseline first.

    We measure the estate from outside and from your own logs, then put an annual cost on the problems we find. Nothing is quoted before this is done.

  2. Blueprint

    Scope and fee agreed in writing.

    The target architecture, the migration path and a fixed fee priced from the survey, itemised so you can see what each part of the work costs.

  3. Build

    Delivered in stages, without downtime.

    Services move across one at a time while the business keeps trading. You review each stage on a live preview, automated tests gate every release, and go-live is a contractual date.

  4. Keep

    Operated under a service level agreement.

    The survey re-runs on a schedule, security patches are applied as they are released, performance is tuned against real load, and an engineer is on call within your agreed hours.

What the survey produces

Findings ranked by severity with the evidence for each, an estimate of the annual cost that you can check line by line, and a recommended order of remediation. The example shown is for a fictional company.

SURVEY DOSSIER · NORTHGATE COLD STORAGE LTD

SAMPLE · FICTIONAL COMPANY · ILLUSTRATIVE FIGURES

DEBT TIER
HIGH
FINDINGS
6
ESTIMATED WASTE
£184,000 / yr
BASIS
Uploaded logs + headcount
SEVERITYFINDINGEVIDENCE
CRITICALPHP 7.2 in productionX-Powered-By: PHP/7.2.34 · out of security support since Nov 2020
HIGHServer responds in 1.6smedian TTFB 1,620ms over 5 samples · target 200ms
HIGHjQuery 1.8.3 on the order portalscript-src /js/jquery-1.8.3.min.js · known published CVEs
MEDIUMBooking API stalls under loadPOST /api/slots p95 2,940ms in the uploaded request log
MEDIUMNo Content Security Policyheader absent on every sampled page
LOWTLS 1.0 still acceptedhandshake completed with TLSv1.0

REMEDIATION ORDER

  1. 01Patch the runtime
  2. 02Cache and index the slot queries
  3. 03Replace the order portal front end
  4. 04Move to the target architecture, service by service

An example of the dossier a survey produces. The company and every figure above are invented. A real dossier is built from your own site, logs and headcount, and the waste figure is an estimate you can check line by line.

Selected work

Platforms we have designed, built and continue to operate, each with its own back office, payments and data.

After go-live

Most of a system’s life happens after launch. Response times are written into your service level agreement, and the controls behind them are published for your procurement team to review.

Standard service levels. Your SLA schedule fixes the hours and any service credits.

P1
Production down or data at risk
First response: 1 hour
Updates: every 2 hours until resolved
P2
Major function impaired, workaround exists
First response: 4 working hours
Updates: daily
P3
Minor issue or change request
First response: 1 working day
Updates: as agreed

Assurance

  • Personal data processed only under a UK GDPR Article 28 agreement, signed before work starts.
  • Client data held in the UK wherever our providers allow it, encrypted in transit and at rest.
  • You own the code, infrastructure and documentation, in accounts you control.
  • Production runs on providers holding independent assurance: Vercel, Neon, GitHub (SOC 2 Type II and ISO/IEC 27001).
Read the trust and procurement page
The engineer who surveys your estate is the engineer who builds it, and the one you call when it matters.

Samuel Deangelo, founder and principal engineer

Tell us about the system that keeps you up at night

Book an infrastructure diagnosis

Send us the domain. We run a first pass on your public systems and take you through the findings on a 30-minute call with the engineer who would lead the work.