AI Engineer

What you will build

Who we are AI ScienTechs is an engineering partner. We introduce vetted Latin American engineers to U.S. companies. This page describes the work. It does not name the client. You learn the company only if you are invited to their interview. About the role You will own the path from a product problem to a reliable AI feature in production. That includes choosing where a model helps, how it is called, how quality is measured, and how failures are handled. You work with product and engineering, not as a research desk detached from the release. What you will do - Turn product use cases into AI features that run in the existing application, with clear inputs, outputs, and failure behavior. - Integrate model APIs and internal services. Own prompts, tool use, retrieval, and the application logic that surrounds them. - Build evaluation sets and review outputs before a change reaches users. Track quality, latency, and cost. - Design retrieval and context so answers stay grounded in the client’s data, with access controls respected. - Add guardrails for privacy, prompt injection, and unsafe output. Document what the system must never do. - Work with backend and frontend engineers so the feature is operable: logs, traces, fallbacks, and a way to turn it off. - Explain tradeoffs in writing and in working sessions with U.S. stakeholders. What we look for - Several years building software that shipped, including at least one production system that called a model or search/retrieval stack. - Strong Python or TypeScript, and comfort reading the surrounding API, data, and frontend code. - Practical judgment about when not to use a model. - Experience with evaluation, not only demos. - Professional English for design reviews and written specs. - A résumé that shows systems you owned, not only courses or notebooks. How you join - The engagement is full-time and remote, from Latin America, with overlap on U.S. business hours. - The working language is professional English: meetings, writing, and review. - Pay is in USD. The amount is agreed with you after evaluation, before any client interview. It is not posted on this page. - We review technical depth and how you work with other people. This is not a marketplace where you bid for the seat. - Create your account and submit your résumé against this role. Your profile stays in our talent network for this opening and for later matches.