Skip to content
QADtechLab

Direction, not a price list

What we're building next — and why we're not selling it yet.

The ownership thesis needs all three layers to make sense, so we publish all three. But we only take money for what we can prove today, and these four are not there yet. Each one below states the gate: the specific thing that has to be true before we would quote it.

The four

  • On-Prem Migration

    Layer 2 · Infrastructure · 3–9 months · Milestones

    Moving workloads out of metered cloud and onto hardware you own, in a colo or your own rack.

    GateOne delivered migration as a reference. Until then we assess and specify, and refer execution to a partner who has done it.

  • Managed On-Prem

    Layer 2 · Infrastructure · Ongoing · Retainer

    Running the hardware after the move — the answer to "who owns 3am?", which is the question that decides most repatriation deals.

    GateA migration reference first, and the on-call capacity to honour it. A retainer we cannot staff is worse than no retainer.

  • AI Readiness Review

    Layer 3 · Intelligence · 2 weeks · Fixed fee

    Which of your workloads should run on your own models, which should stay on an API, and which should not use a model at all.

    GateProfessional indemnity cover in force, since this is advisory work in a domain where being wrong is expensive for the client.

  • Private AI Stack

    Layer 3 · Intelligence · 6–12 weeks · Milestones

    Open-weights models on your own GPUs, for the narrow high-volume work where that genuinely beats an API — classification, extraction, tagging, routing, embeddings.

    GateA delivered readiness review, and honesty about the split: hard reasoning and long-context work still belongs on a frontier API, and pretending otherwise is discovered in week two.

Why publish this at all