FAQ
Frequently Asked Questions
Straight answers to the questions we are asked most often. If yours is not here, ask us directly — we would rather answer it properly.
Consultation
Illustrative- YouWe get about a hundred inbound enquiries a month, all coming in by email. Right now someone re-types each one into the CRM by hand.09:12
- DevRoxThat is a classic discovery candidate. First step would be mapping how a lead actually moves through your team today — then we flag which steps an automated handoff can take over.09:13
- YouWould it be able to draft a reply for us, or does it need a human on every message?09:15
- DevRoxBoth patterns exist. The honest answer is we would not know where yours lands until we see the enquiry mix — so the first deliverable is a scoping note, not a demo.09:16
Engagement
How does an engagement usually start?
With a 30-minute consultation. We walk through the problem, the systems already in place and what a good outcome would look like. If it is a fit, the next step is a short paid discovery that produces an architecture, a scope and a fixed delivery plan before any build work is committed.
Do you take on small projects?
Yes. A single well-chosen automation often pays for itself faster than a large platform. If the scope is genuinely small we will say so and price it accordingly rather than inflating it into a programme of work.
Who will we actually be working with?
The engineers building your system, directly. There is no layer of account managers relaying information, and the people in the discovery sessions are the people who write the code.
Delivery
How long does a typical project take?
A focused automation or integration is usually two to five weeks. A voice agent or AI assistant with real system access is typically six to ten weeks. A full platform build runs three months or more, delivered in increments so you have working software throughout rather than at the end.
What happens after launch?
Every system ships with monitoring and alerting already in place. Most clients continue with a support arrangement covering incident response, dependency updates and a monthly improvement allocation. Some take the system fully in-house after handover, which is a perfectly good outcome.
Technology
How do you decide where AI is actually appropriate?
During discovery we separate the steps that need judgement from the steps that just need reliable logic. AI goes where judgement is required — classification, extraction, summarisation, conversation. Everything else stays deterministic, because it is cheaper, faster and far easier to audit. It is common for a project to end up using less AI than expected.
Can you work with our existing systems?
Yes, and it is the usual case. We integrate with CRMs, ERPs, databases, telephony and internal APIs rather than asking you to replace them. Where a legacy system has no usable API we work through the interfaces that do exist and isolate that dependency behind a clean boundary.
How is our data handled when AI models are involved?
Data handling is agreed in writing during discovery. We default to sending the minimum necessary to a model, redacting identifiers where the task allows it, and using providers whose terms exclude training on your data. Where requirements are stricter we use self-hosted or open models instead.
Commercial
Who owns the code and the infrastructure?
You do, from the first commit. Work lands in a repository you control, infrastructure runs in your accounts, and handover includes architecture documentation and a walkthrough for your engineers. There is no proprietary runtime you have to keep paying us for.
How do you price work?
Discovery is a fixed fee. Delivery is either fixed-scope for well-defined builds or a monthly retainer for ongoing product work. You get the estimate and the assumptions behind it in writing, and we flag scope changes when they happen rather than at invoicing.
Tools we build with
Tools we build with
- React
- Next.js
- TypeScript
- Tailwind CSS
- Node.js
- Express
- Python
- FastAPI
- PostgreSQL
- MongoDB
- MySQL
- Firebase
- OpenAI
- Gemini
- Claude
- Hugging Face
- n8n
- Make
- GoHighLevel
- Zapier
- Twilio
- ElevenLabs
- Vercel
- Docker
- GitHub Actions
Still deciding?
Have a Question We Have Not Answered?
Book a free consultation or send us a message. There is no obligation and no sales script.
hello@devrox.comWe reply to every enquiry within one business day.