Zynterra

Services

We would rather be excellent at four things than average at ten. Every engagement falls into one of these, with the same honesty about when it's the right call and when it isn't.

01

Product engineering

Web and mobile applications, built end to end, from the first wireframe to something real people use. Clean architecture, typed code, and a codebase the next engineer can actually read.

Timeline
First working version typically ships in 6 to 12 weeks.

What you get

  • 01A running staging environment you can click through
  • 02A typed, tested codebase with CI on every push
  • 03Written architecture decision records for anything expensive to reverse
  • 04Full handover documentation, not a wiki nobody maintains
  • 05A recorded walkthrough of the codebase for your team

Right call when

  • You have a product to build and no engineering team yet
  • You need a second team to ship a defined feature fast
  • Your MVP proved the idea and now needs to be built properly

Not the right call when

  • You need one contractor embedded long-term in an existing team. That's a retainer, not a project
  • The scope isn't defined yet. Start with a discovery conversation first
02

Cloud & platform

The infrastructure a product depends on and rarely gets attention until it breaks: servers, databases, deployment pipelines, monitoring, and backups that have actually been tested.

Timeline
A migration or setup typically takes 2 to 6 weeks; ongoing run is a retainer.

What you get

  • 01Infrastructure as code, not a server someone clicked together once
  • 02Monitoring and alerting configured before launch, not after the first outage
  • 03Documented runbooks for the failures we can predict
  • 04Backup and restore tested end to end, not just scheduled
  • 05A written incident response plan

Right call when

  • Your app works but goes down under load
  • You're moving off a provider you've outgrown, or one that's gotten too expensive
  • Nobody can say with confidence how a restore from backup would actually go

Not the right call when

  • You need 24/7 on-call coverage starting today. We can build toward that, but it isn't a day-one deliverable
03

Applied AI

Machine intelligence used where it actually earns its keep: search, automation, document processing, decision support. Not AI for the sake of a pitch deck line.

Timeline
A validated prototype typically takes 2 to 4 weeks, before any production build.

What you get

  • 01A working prototype, validated against your real data before we build the full version
  • 02Integration with the systems you already run, not a standalone demo
  • 03Cost monitoring on model usage, so the bill doesn't surprise you
  • 04A defined fallback path for when the model gets it wrong

Right call when

  • You have a specific, well-defined task a model can meaningfully improve
  • You're spending real staff hours on something pattern-based that could be automated

Not the right call when

  • You want "AI" as a feature for a pitch deck with no specific problem behind it. We'll ask what problem before we build anything
04

Security & consulting

Security reviews, hardening, and honest technical advice, plus training so your team can run what we hand over. You're never locked into us for the answers.

Timeline
A focused review typically takes 1 to 3 weeks depending on surface area.

What you get

  • 01A written report naming specific findings, not a generic checklist
  • 02Findings prioritised by real risk, not by how alarming they sound
  • 03A remediation plan, not just a list of problems
  • 04A follow-up review once fixes are in

Right call when

  • You handle payments or personal data and have never had a review
  • You're about to raise or get acquired and need a clean technical picture

Not the right call when

  • You want a compliance checkbox exercise. We'll do the review honestly, which means we won't just say "looks fine" to make a deadline

Approach

How a project runs.

The same four stages, every time. Predictable for you, disciplined for us.

01 · Scope

Agree the ground

A short discovery sprint. We define what's being built, what it costs, and what "done" means, in writing, before any code.

02 · Build

Ship weekly

Working software every week, not a big reveal at the end. You see progress in the product, not in slide decks.

03 · Launch

Land it clean

Tested, documented, and deployed to infrastructure you own. Handover includes everything: code, credentials, and know-how.

04 · Run

Keep it standing

Monitoring, maintenance, and support after launch. The name is the promise: what we build stays on solid ground.

Engagement models

Three ways to work with us.

Whichever fits, the number is agreed in writing before work starts.

Best for a defined deliverable

Fixed-scope project

You know roughly what you want built. We scope it in writing, agree a fixed price before work starts, and price any change before it happens, not after.

  • One number, agreed up front
  • Scope changes priced before they happen
  • Best for a first version or a bounded feature

Best for ongoing build and run

Monthly retainer

A standing team on your product: new features, fixes, and the unglamorous work of keeping something alive after launch. Capacity-based, billed monthly.

  • Predictable monthly cost
  • Same team, so context isn't rebuilt every sprint
  • Pause or scale as the roadmap changes

Best for an existing codebase

Technical review

A paid, fixed-fee assessment of what you already have. You get a written report either way, whether or not you hire us for what comes next.

  • Written report, not a sales pitch
  • A stabilise, improve, or rebuild recommendation
  • Sometimes the honest answer is "keep it"

How pricing works

  • 01Scope is agreed in writing before any work starts
  • 02Changes to scope are priced before they happen, never after the fact
  • 03No hourly billing surprises. You know the shape of the number going in
  • 04Invoicing is tied to milestones, not to time elapsed

What we need from you

  • 01One decision-maker who can approve scope and sign off on progress
  • 02Timely feedback. A project waiting on a reply loses its own momentum
  • 03Access to the accounts and systems the work actually touches
  • 04A willingness to hear it when something is a bad idea. That's most of what you're paying for

Tools & stack

Defaults, not limits.

What follows is where we start, not where we stop. We pick a default so that decisions are fast and boring, then work in whatever your product already runs on.

GoJava & Kotlin.NET / C#PHP & LaravelRuby on RailsRustFlutterSwiftDjangoGraphQLAWSGoogle CloudAzureTerraform

We also work in — ask about anything not listed

Frontend

  • Next.js
  • React
  • React Native
  • TypeScript
  • TailwindCSS

Backend

  • NestJS
  • Fastify
  • Node.js
  • Prisma
  • Python

Data & infrastructure

  • PostgreSQL
  • Redis
  • Docker
  • Kubernetes
  • Hetzner

Delivery & operations

  • Cloudflare
  • GitHub
  • Playwright
  • OpenTelemetry

Our defaults — the stack Zorva and AuditGate run on

We start from what we run ourselves

The defaults above are the stack Zorva and AuditGate run on. We recommend them because we carry the pager for them, not because they are the only things we know.

Your existing stack is not a problem

Takeover work usually lands in a language we did not choose. If your team already runs Go, Java, .NET, PHP, or something older, we work in it rather than quietly rewriting it.

The standards travel, the tools change

Typed interfaces, tests on anything touching money or auth, review before merge, monitoring before launch. None of that belongs to a language. It is the part we will not drop whatever we are writing in.

Read the engineering standards

Principles

What we won't compromise.

P-1

Build for month six

Most software doesn't fail at launch. It fails later, when nobody's watching. We design for the day the demo is forgotten and the system just has to work.

P-2

You own everything

Your code, your cloud accounts, your data. We hand over the keys as we go, and we document as we build. Lock-in is a business model. It's not ours.

P-3

Say the true thing

If a feature isn't worth building, we'll say so. If a deadline is at risk, you'll hear it from us first. Plain answers cost less than pleasant surprises.

Start a project
Read how we work