Questions

Nothing. It takes 30 minutes, and you leave with our opinion on your architecture or AI approach, whether or not we work together.

Yes. Send it when you book, and we sign it before we talk.

What you are trying to build, what exists today, and any deadline or budget you are working to. Slides are not needed.

Yes. We explain every decision in plain words, and the weekly written update is written for the person who decides, not only for engineers.

The prototype in your repository, the eval set with its numbers, and a written go/no-go report.

You keep the prototype, the eval numbers and a written reason, two weeks in.

That is normal, and the sprint will show it. The report says what the data needs before production.

A sample of real data, one domain expert to agree the eval set, access to a sandbox or permission for us to host it, and the decision maker at the day 10 review.

The senior engineer who leads the project and writes its code. When a collaborator joins, you meet them before they start.

Yes: your repository, your issue tracker, and your Slack or Teams.

We keep working hours that overlap with the US, Europe or Australia, agreed per project.

Yes. Embedded Engineering puts us inside your team by the month, in your stand-ups and your code review.

We write the change down with its cost and time before any work on it starts. Nothing changes silently.

Yes. Every step ends in documents you keep, and none of them commits you to the next.

Yes. The first week is a written review of what is there: what to keep, what to fix first, and what it would cost.

Not to zero. We measure how often it happens on your eval set, make it cite sources, and make it refuse when it is unsure. You see the rate before launch.

Usually RAG first: it is cheaper to change and every answer can show its source. Fine-tuning is worth it when the eval set shows RAG cannot reach the target.

The one that meets your targets for the lowest cost on your eval set. We test more than one.

The PoC Sprint measures cost per answer on your data, and the go/no-go report turns that into a monthly estimate for your volume.

Yes. We add RAG, copilots and document automation to the apps and databases you have today.

You do. All code and IP from the project belong to you.

Only if you choose a provider that way. Models can also run through your cloud account or on your own servers.

Yes. Code, data, models and logs can all run in your account, so nothing leaves it.

No. We are a small studio, so we work inside your controls instead: your cloud, your accounts, your policies, and your vendor review.

Yes, for a scoped build: the estimate from Discovery & Architecture becomes the price. Changes are written down with their cost before work starts.

$2,000, fixed, agreed before day 1. Every price is on the pricing page.

Your team, with the docs and runbook from the handover, or us, through Embedded Engineering by the month.