// services
Back End &
Infrastructure
The front end gets the compliments; the back end keeps the promises. We build and run the server side - APIs, databases, email, and the business logic in between - so the whole thing holds up under real use.
What is included: Back End & Infrastructure
APIs & services
REST and edge APIs with the resilience layer done properly: retries, timeouts, and circuit breakers - the same primitives we ship in our own open-source QoS package.
Databases
Schema design, migrations, and query work on Postgres, SQLite, and managed platforms - built for the access patterns you actually have.
Email infrastructure
Transactional email, routing, and deliverability - configured so order confirmations and password resets do not vanish into spam folders.
Hosting & deploys
Cloudflare Pages, Workers, and edge deployment pipelines with caching, redirects, and monitoring set up from the start.
The parts nobody sees, and why they decide everything
Users judge a product by the front end, but they experience the back end. It is the back end that decides whether an order confirmation arrives, whether the site stays up on your busiest day, and whether a flaky third-party service takes you down with it.
That is why we treat infrastructure as product work rather than a hosting afterthought. The visible features are the last ten percent; the boring plumbing is what makes them reliable.
Resilience is a design decision, not a setting
Most outages in small systems come from one place talking to another and one of them being slow. A payment provider has a bad afternoon, a request waits forever, and the whole site stops responding.
We build against that pattern by default: hard timeouts so nothing waits indefinitely, retries with backoff so a blip does not become an error, and circuit breakers so a failing dependency degrades gracefully instead of cascading. These are the same primitives we publish as our own open-source QoS package, so the approach is not theoretical.
Data you can trust and move
Schema design is where small projects quietly go wrong. Tables built for the first version of an idea become painful at the second, and migrations on a live database with real customer data are the least forgiving work in software.
We design for the access patterns you actually have, write migrations that can be run safely on production, and leave a clear record of what changed. When the time comes to move or split the data, that record is what makes it possible.
Hosting you understand
You should not need a specialist to answer "is my site up." We set up edge hosting with caching, redirects, health checks, and monitoring that tells someone when something breaks - ideally before a customer does.
Accounts are in your name, costs are transparent, and there is no proprietary layer you cannot leave. If you ever want to move, the exit is documented rather than discovered.
Common questions
Can you take over our existing back end?
Yes. We start with an audit of what is there, then stabilize before we change anything - inheriting a mess is our specialty, not an exception.
Do you handle the hosting bill and accounts?
We set everything up in accounts you own. You keep the keys; we do the work. No lock-in, no hostage infrastructure.
Our site goes down when traffic spikes. Can you fix that?
Usually, and usually without rebuilding everything. Caching, query work, and moving static assets to an edge network solve a surprising share of load problems. We diagnose before we prescribe.
Do you do server administration?
For managed and edge platforms, yes as part of the build. For traditional servers we handle it under our IT contracting service, which is the better fit for ongoing operations.
Can you work with our database as-is?
Yes. Most engagements improve an existing schema incrementally rather than starting over. A rewrite is a last resort, not a default, because the risk sits with your data.