Common questions
Before we build together.
Practical answers about working with NorthBlanc, from project scope to ownership and support.
Questions
A little more clarity.
Good work starts with good questions. These are the ones we hear most often.
What can we build together?
NorthBlanc is an AI-native software development company. We build AI assistants, LLM applications, retrieval systems, web platforms, and cloud infrastructure. You can engage us for a complete project or bring our senior engineers into your existing team.
How soon will we see something working?
Scoping takes about a week, followed by a two-to-three-week prototype focused on your hardest problem. Most full projects reach production in eight to sixteen weeks from kickoff, with working software to review every two weeks.
How do you price a project?
Scoping has a fixed fee. Once the scope is clear, choose a fixed-price build against an agreed specification or a monthly rate for an embedded team. We provide a quote after understanding your needs and data.
Can your engineers join our team?
Yes. Our engineers work in your repository, join your standups, and collaborate with your leads. Embedded teams are billed monthly, with thirty days’ notice on either side and no minimum term.
Do you work internationally?
Yes. We work with clients in North America, the UK, Europe, and the Gulf. We keep a four-hour overlap with European and UK working hours and a shorter overlap with US teams. Invoices can be in USD, EUR, or GBP.
What if our data or existing code needs work?
That is a good place to start. We assess your data, code, tests, and running costs before proposing changes. You get a clear account of what to keep, what to improve, and what needs to happen first.
Who owns the finished product?
You do, from day one. Code, designs, prompts, evaluation sets, and infrastructure definitions belong in your accounts and repositories. We document the work and plan a handover or ongoing support around your team.
Are we tied to a single AI provider?
No. We evaluate models against your tasks, quality requirements, and budget, and keep the model layer replaceable. The provider is a decision we make with evidence, and revisit when your needs change.