Three ways to start.

Each is a different size of commitment, and they are meant to be read as a ladder — the audit answers whether the rest is worth doing.

  1. LLM feature audit

    Two weeks, fixed price, a written deliverable

    A read of the layer your AI features actually run on. Is there one registry of what the system can do, or several? Is there a single admission point, or does every caller check authorisation itself? Are host-side concerns separated from the schema the model sees?

    It ends in a document you keep — findings, the decisions behind them, and what each one will cost to change. If the answer is that nothing needs doing, the document says that.

  2. Fixed-scope build

    Prototype through to production

    A defined piece of work with an agreed boundary: an integration surface, a tool layer, a migration off a prototype that outgrew itself. Scoped before it starts, and scoped narrowly enough that the estimate means something.

    Most of what makes an AI feature survive contact with production is decided before any model is called. That groundwork is the work.

  3. Fractional engineering

    Ongoing, days per month

    Continuous involvement with a team that has the capability but not the capacity — architecture review, the difficult parts, and the judgement calls that are expensive to get wrong and cheap to get right early.

    Suited to a team shipping steadily who want a second opinion that has seen the failure modes.

What we don't do

Staff augmentation — a developer placed into your team under your management, billed by the day. It is not what this is for, and it is the one shape where nobody involved gets the benefit of the engineering judgement they are paying for.

We also don't publish day rates. Price depends on scope and on what the first conversation turns up, and a number posted before that conversation anchors it in the wrong place for both sides.

Starting

Email hello@moltera.dev with what you are building and where it is getting difficult. You will get a direct answer about whether this is work we can do well, including when it isn't.