Questions that usually come up before a quote.

After a short discovery we put together scope, phases, and investment in MXN or USD, whichever suits you. Catalog ranges are a reference; the final number is confirmed once the problem is bounded. There is no serious quote without that conversation.

Yes. We are in Mérida and meet in person on the peninsula when it makes sense. The rest of the work is remote, with demos and a clear decision channel.

No. AI is a strong line — research and product — but we also build custom software, data science, and sites / landings. Sometimes the right answer is not a model.

That is the most common case. We design the AI piece (RAG, agent, vision, scoring) and connect it to your APIs, documents, or channels, with usage limits and a human handoff.

A campaign landing usually takes 2 to 4 weeks; a corporate site, 4 to 8, depending on content and integrations. We confirm it in the proposal, not as a generic average.

Whatever the problem asks for. Web product: Next.js, Tailwind, Vercel. Backend: Node/NestJS or Python/FastAPI. AI: PyTorch, RAG, vision, agents. Cloud according to your ecosystem (AWS, GCP, or Azure).

Confidential handling of data, access, and code. Ownership of client work is set by contract; we do not reuse sensitive information across accounts.

Yes. After release there is stabilization and, if you hire it, a maintenance plan aligned with the product lifecycle.