Senior AI Engineer

ABOUT VYSTA

Vysta Paid Media Group is a Google Premier Partner performance marketing agency managing more than $250 million in annual Google and YouTube ad spend across 150+ ecommerce brands.

We are building AI-powered systems that help our team turn performance data, client knowledge, and operational workflows into better decisions and more efficient execution. Our focus is practical: reliable tools that people use, measurable improvements in how we work, and appropriate human oversight.

THE ROLE

We are looking for a hands-on Senior AI Engineer to work directly under our Head of AI and turn our AI roadmap into production-ready applications, agents, and automations.

You will take ownership of individual builds from technical design and implementation through testing, deployment, monitoring, and ongoing improvement. Working with our Head of AI and the teams using these systems, you will solve real operational problems and make sure the solutions remain useful and reliable after launch.

This is an applied AI engineering role. We need someone who can write and maintain production software, integrate models with tools and APIs, and manage the failure modes of real-world AI systems. Building an impressive demo is a starting point, not the finish line.

WHAT YOU WILL DO

• Build production AI agents and LLM-powered applications that retrieve information, use tools and APIs, and complete multi-step workflows.

• Connect AI systems to business data and tools, including advertising platforms, analytics sources, internal knowledge, and communication or task-management systems.

• Own the code behind your builds: repositories, code reviews, automated tests, deployment pipelines, documentation, monitoring, and maintenance.

• Engineer for reliability: validate inputs and outputs, handle API failures and rate limits, implement retries and fallbacks, and make failures visible and recoverable.

• Evaluate and improve quality: create representative test cases, measure accuracy and task completion, and catch regressions when prompts, models, or integrations change.

• Build appropriate controls: protect client data, enforce account boundaries and permissions, and require human approval before consequential actions such as material campaign changes or external client communications.

• Use AI coding assistants effectively, while reviewing, testing, and taking responsibility for the code they produce.

• Work closely with our Head of AI and operational teams to scope projects, iterate on feedback, support adoption, and measure results. Use deterministic automation instead of an LLM when it is the better solution.

EXAMPLES OF WHAT YOU COULD BUILD

• An account-review assistant that gathers performance evidence, flags missing or conflicting data, and prepares a review for a media buyer.

• A reporting workflow that produces evidence-backed drafts with client-specific data checks and approval before sending.

• A monitoring workflow that turns meaningful performance or tracking issues into assigned tasks with clear follow-through.

• An internal knowledge assistant that retrieves approved processes and account context, with source references and appropriate access controls.

REQUIRED EXPERIENCE

• At least 5 years of professional experience building AI systems, with hands-on responsibility for shipping and maintaining production software. Earlier machine-learning experience can count; this does not mean five years using today's LLM tools.

• At least 3 AI agents or LLM-powered applications that you personally built and deployed to production. You should be able to explain your contribution, the architecture, how the application was used, and how you measured its performance.

• End-to-end ownership of a production AI codebase, including its repository, tests, deployment, and monitoring.

• Strong software engineering skills in Python and/or TypeScript, with practical experience working with APIs, databases, and SQL.

• Hands-on experience with LLM APIs, tool/function calling, structured outputs, retrieval, and multi-step workflow orchestration.

• Experience deploying an agent that uses tools or APIs, handles errors, and includes human approval for consequential actions.

• Regular use of at least one AI coding assistant, such as Cursor, Claude Code, or Codex, to build production software.

• Experience deploying and operating applications in a cloud environment, with sound practices for credentials, permissions, logging, and monitoring.

• Clear communication, strong ownership, and the ability to work effectively with technical and non-technical teammates in a remote environment.

HELPFUL, BUT NOT REQUIRED

• Experience in ecommerce, performance marketing, ad tech, or agency operations.

• Familiarity with Google Ads, YouTube advertising, analytics, or related APIs.

• Experience with retrieval-augmented generation, agent frameworks, or workflow tools such as n8n or Make, alongside strong coding skills.

• Experience measuring AI quality, latency, cost, and reliability in production.

A relevant degree is welcome, but demonstrated engineering ability and production experience matter more. Custom model training or neural-network research is not a prerequisite for this role.

WHAT SUCCESS LOOKS LIKE

• The systems you build are used consistently by the team, not just demonstrated once.

• Your applications are tested, documented, observable, and maintainable by other engineers.

• Improvements are measured through time saved, fewer errors, better task completion, or faster access to reliable information.

• Client boundaries and approval controls hold up in real-world use.

• You deliver against priorities agreed with our Head of AI and communicate progress, risks, and trade-offs clearly.

WHY JOIN VYSTA

You will build AI systems around real workflows and substantial advertising operations, working directly with the people who use your software. You will have hands-on ownership of meaningful projects, guidance from our Head of AI, and a clear connection between your engineering work and business outcomes.