AI Automation Engineer, In-House MarTech Video SaaS

Headquarters: Austin, TX URL: https://vidalytics.com If you love building AI agents, wiring systems together, and watching whole teams stop doing by hand what a machine can do better, faster, and cheaper, and you want to make a whole company AI native rather than ship a slick demo, keep reading. You'll be the person who makes Vidalytics AI native from top to bottom. You'll build the agents, the chains, the prompts, the reusable skills, and the cross system automations that let every team punch way above its weight. You'll report to the CEO. LFG! You'll start where the money is, on the revenue teams (sales, marketing and customer success), then take it across the whole org. You’ll work with the stakeholders who will use your workflow automations on a daily basis, to achieve the best solutions. Most importantly work w/ each team member to help them become AI Native themselves. Which tools they should be using, how they can build upon your ideas, or even spearheading projects of their own that you ensure the success of from a technical perspective. What You'll Be Doing (Job Tasks) Build AI agents and chains that do real work end to end, not chatbots that just talk. Write, version, and test prompts like code. Build reusable skills the whole org can run. Stand up cross system automations (n8n, Zapier, and custom code) that wire our tools together so work flows without a human babysitting it. Make each team at Vidalytics AI native,. Not just automating work, but helping each team member become AI Native - knowing which tools to use and how to use them. Where they’re self sufficient to continue automating their jobs. Ship internal tools and automations that make the company more efficient and effective. You stay hungry for more, always improving what we build and use. Build evals and observability so every AI system is measured, monitored, traced, and cost controlled. If you can't measure it, you can't ship it. Ground AI outputs in our own company data (retrieval / RAG) so they're accurate, not confidently wrong. Diagnose failures, kill hallucinations, and build the guardrails that keep systems reliable in production. Help with AEO / GEO where it moves the needle. Requirements (The must haves) You’re able to choose the right tool for the job. Which AI tool, LLM, agent is the best answer you can figure out wihtout endless tinkering. You care whether the work gets faster, cheaper, or better, not whether it uses the trendy thing. You've already shipped this kind of work in production. This is not a learn on the job role nor is it for the hobbiest. We expect real results, right away. We’re already doing all of this ourselves but need an expert level person to take ownership. AI is a young field, so we expect only a few years of hands on AI experience, but it has to sit on top of strong, relevant broader engineering experience, and you must have built and run real AI systems that other people depend on. You can actually code, and not just with Claude lol. You're comfortable with databases, APIs, and MCP, wiring systems together for real, not just clicking boxes. You build evals and observability by default. You measure, monitor, trace, and control cost as a matter of habit. You hold a production grade reliability bar. You ship systems that keep working for months, not demos that impress once and break the next day. A ready to show portfolio, or concrete examples of what you've built and the results it delivered. Your working hours overlap strongly with US Eastern time. We can discuss further. You’re not token-wise, prompt foolish. You're fluent across the frontier models, (Claude is our primary, ideally it is yours), and you know which model to reach for and why (cost, latency, context, reliability, specific use cases, etc.). You use tools like Claude Code and Claude Cowork daily, and not just a prompt You're comfortable with AI native hosting and deployment, for example Vercel. You're hands on with automation tools like n8n and Zapier, a