How do engagements usually start?
A 20-minute intro call to understand the problem, followed by a paid 1-week discovery sprint where we produce a written recommendation and a fixed quote. You decide whether to proceed.
How do you price?
We quote fixed prices per sprint based on the scope agreed at the end of discovery. Sprints typically run 4–8 weeks. Once you're live, ongoing support is billed on a small monthly retainer or per-incident.
Who owns the code and the data?
You do — from day one. Source is committed to a repo you own, infrastructure runs in your cloud account, and all data stays inside your environment. We document handover so any competent engineer can pick it up.
Do you work with our existing team or replace it?
We almost always work alongside existing teams. We're a force multiplier — your people know the business, we bring the engineering practices.
Where does the data live? Is it POPIA-compliant?
By default, data lives in your own cloud account in a region you choose — typically AWS af-south-1 (Cape Town) or Azure South Africa North for POPIA-sensitive workloads. We build with POPIA principles (data minimisation, purpose limitation, access logging) baked in.
How fast can you start?
Usually 1–3 weeks from signed agreement. We deliberately keep a small client base so we can give each engagement real attention.
Do you do training-from-scratch AI work, or only fine-tuning?
Fine-tuning and retrieval-augmented generation (RAG) cover ~95% of business cases more cost-effectively than training from scratch. If a foundation-model approach is genuinely the right answer, we'll tell you and help you scope it accordingly.
Will you sign an NDA?
Yes, happily — either yours or ours. Mutual NDAs are usually signed before the discovery sprint begins.