Frequently asked questions

Most focused first releases land between USD 15,000 and USD 60,000 depending on scope, integrations and compliance requirements. Websites typically run USD 5,000–20,000, a single automation or integration USD 5,000–20,000, and a first ERP module USD 25,000–60,000. You get a fixed price for a defined phase after a discovery call.

Fixed price per phase. We scope a phase tightly, quote it, and hold that price. If scope changes mid-phase we raise a written change request rather than quietly billing more.

Yes, typically an initial milestone payment to begin, then milestone-based billing through delivery. Payment terms are set out in the service agreement before work starts.

No. You own what we build, so there is no per-seat or per-year licence to keep using it. Your only ongoing costs are hosting, any third-party services you choose, and optional support.

We prefer it. A small first phase that solves one expensive problem is a better test than any sales process, and it means both sides know what they are committing to before the larger work.

Websites typically launch in 4–8 weeks. A useful first release of custom software ships in 8–16 weeks. A first ERP module runs 10–14 weeks. Integrations and automations are usually 3–8 weeks.

Usually within one to three weeks of agreement, depending on current commitments. We will tell you our real availability rather than promising an immediate start and then queuing you.

In our experience, almost always one of three things: delayed client feedback or approvals, third-party dependencies that behave differently than documented, or scope added mid-phase. We flag all three early and track them explicitly.

Yes. You get working previews at every milestone, not a status report. If something is drifting you will see it in the software long before a deadline.

Yes. On full payment for a milestone you own the source code and custom assets outright, delivered in your own repository. There is no lock-in.

We retain ownership of our pre-existing internal tooling, libraries and general know-how, and may reuse non-confidential, non-client-specific components elsewhere. This never restricts your ownership of your system.

Yes, and we write code on the assumption that you might. Standard frameworks, documented setup, and no proprietary layer you would need us to maintain.

You do, always. We process it to deliver the engagement and nothing else, under a signed agreement.

Every build includes a warranty period covering defects against the agreed scope. Beyond that, most clients keep a support and improvement retainer, because the highest-value changes are the ones you discover once real users are in the system.

Retainers are sized to the system and the response times you need. Small systems often need only a few hours a month; larger operational platforms need more. We size it from actual usage rather than a standard package.

No. You own the code and can maintain it yourself or with another team. A retainer is an option, not a condition of the build.

Within one business day for standard requests. Retainer clients with operational systems get agreed response targets for critical issues, defined in the retainer.

Primarily TypeScript across the stack — Next.js and React on the front end, Node.js on the back end, PostgreSQL or MongoDB for data, hosted on AWS, Vercel or Cloudflare. We use Python where the work is data or ML heavy.

They are mainstream, well-documented and widely staffed, which matters more than novelty. You should be able to hire for or hand over your system without hunting for a specialist in something obscure.

Usually yes. Where you already have a working platform we extend or integrate with it rather than proposing a rewrite for our own convenience.

Yes, typically React Native for cross-platform, or a well-built responsive web app where that genuinely serves the use case better and costs less to maintain.

No. We use enterprise API tiers where your data is not used for model training, and can architect for data residency or self-hosted models where regulation or client contracts require it.

We design for that. Responses are grounded in your own approved content rather than general model knowledge, low-confidence cases route to a human, and anything consequential sits behind human approval by default.

No, and we will tell you when it is not. A well-defined rule or a simple integration is often more reliable and far cheaper than a language model. We recommend the boring option when the boring option wins.

We agree the metric before the pilot — usually hours saved, first-response time, or documents processed per hour — and instrument the system to report it. If the pilot misses the threshold, we say so.

Encryption in transit and at rest, role-based access control with authorisation checks on every request, audit logging, dependency scanning, and secrets kept out of source control. Security requirements are scoped during discovery rather than retrofitted.

We design to the relevant controls — access control, encryption, audit trails, retention and data residency. We are engineers rather than your compliance auditor, so we build to the requirements your advisors define.

In the region you require. We commonly deploy to US, EU, UK, Australian or Indian regions depending on client and regulatory needs.

We run security review and dependency scanning as standard. For systems handling sensitive data we recommend an independent penetration test, which we can coordinate and then remediate against.

A 30-minute discovery call, then a business audit that maps your workflows and quantifies the opportunity. Only after that do we propose scope, cost and sequence — so the proposal is based on evidence rather than assumptions.

A single empowered point of contact, timely feedback on previews, access to the systems in scope, and honesty about how work really happens — including the workarounds.

Through a written change request estimating effort, cost and timeline impact, approved before work begins. No silent scope creep in either direction.

Milestone-based delivery means you are reviewing working software throughout, so problems surface early rather than at the end. If a milestone does not meet the agreed scope, we fix it before it is billed as complete.

Usually yes. We routinely integrate accounting platforms, CRMs, ERPs, payment providers, email, telematics and industry-specific systems via API.

There are still options: scheduled file import and export, database-level integration, vendor middleware, or a controlled migration. We assess this during the audit before promising anything.

We design for it — queued retries, idempotent writes so nothing double-posts, alerting when something stalls, and a documented reconciliation path. An outage becomes a delay rather than a data-integrity incident.

Usually you do, in your own cloud account, so you retain full control and the billing relationship. We set it up, document it, and can manage it under a retainer if you prefer.

For most small and mid-sized systems, typically USD 20–200 per month depending on traffic, data volume and redundancy requirements. We estimate it during solution design so there are no surprises.

Yes. Automated backups, uptime and error monitoring, and alerting are part of a production launch rather than an upsell.