Performance optimisation
Your system, faster, with the before-and-after to prove it.
Load times, database queries, front-end bundles and the servers underneath. We measure first, then fix where the measurement says the time goes, and leave monitoring behind so it does not creep back.
Bring the system that is too slow.
- A working session on your live system and its real timings.
- Written findings: where the time goes and what would fix it, in order of payoff.
- Yours to act on with us or with anyone else.
- 12years building
- 43systems delivered
- 37organisations
- 18sectors
What you get
Six places the time goes, and what we do about each.
Every fix is measured before and after, in your own repository, so you can see what changed and by how much.
Pages that arrive before the user gives up
Code splitting, lazy loading and asset optimisation, applied where your own timings show the wait is, rather than everywhere at once.
Content served from near your users
A content delivery network in front of your static assets and media, with caching rules your release process can live with, for users wherever they are.
Queries that stop being the bottleneck
Slow queries found from your real workload, then indexing, query rewrites and caching, with the query plans from before and after so you can see the gain.
Less JavaScript, better images, smoother rendering
Bundle analysis and trimming, modern image formats and sizes, and rendering fixes for the stutter users notice on mid-range phones.
Capacity that grows with a traffic spike
Auto-scaling infrastructure sized to your real peaks (an event day, a results release, a campaign) and back down when they pass, so you pay for capacity only while you need it.
Alerts when it degrades
Real-time performance monitoring with alerts on degradation, tied to the metrics that matter to you, so a regression becomes a ticket rather than a complaint.
Work in this area
Platforms, an API and an event-day voucher system we have delivered.
8 of the 43: three self-service builder platforms, an OCR API, and a redemption system for a Chinese New Year event. These are the kinds of systems where response time and load are the job.
All 43 systems
In writing
Three things you get before the first invoice.
01
The measurements are yours.
We hand over every profile, benchmark and before-and-after figure as it is produced, in your repository and your monitoring, rather than holding it back until the final invoice.
02
Phase one is fixed price.
The audit of where the time goes is a defined scope at a defined cost, and the report is yours whether or not you continue with us.
03
A warranty, in writing.
Defects in work we shipped, we fix at no charge for the warranty period. A support agreement with named response times is available after that.
How it runs
Five stages. You receive something you own at each.
- 01Insight
A written profile of where the time goes, taken from your real traffic.
- 02Strategy
The fix list in order of payoff, with a fixed price for phase one.
- 03Build
Each fix landing in your repository with its before-and-after measurement.
- 04Launch
The production release and the monitoring that watches it.
- 05Support
Documentation, a warranty, and named response times if you want them.
Other services
Performance work often starts from an architecture question.
Start with a free consultation on your system.
Bring the system your users complain about, or the launch you expect to be busy. We spend a working session on it and you leave with written findings that are yours to act on, with us or with anyone else.