Marketing Engineer

Taxfyle•Miami, FL

5d•$85,000 - $95,000•Onsite

About The Position

The Marketing Engineer is the applied-AI builder inside our marketing team. This person ships the AI systems that do marketing work: agents that research and draft, pipelines that produce content at scale, and automations that carry a lead from first touch to close without anyone copying data between tools. This is a hands-on engineering role, not a prompt-writing role. Our website, HubSpot instance, and analytics stack are the substrate the AI runs on, so this role owns them too. The mandate is to compound the team's output through systems. Work is done through version control, peer review, and staged deployment, in close partnership with Marketing, Design, Sales, Engineering, and Product.

Requirements

Professional experience in web development, marketing engineering, or marketing operations, with strong JavaScript/TypeScript, HTML, and CSS. React, Next.js, and Node.js required.
Demonstrated experience building with LLMs beyond chat use: API integration, prompt engineering, tool calling, structured outputs, and retrieval-augmented generation. A portfolio of shipped AI workflows matters more than credentials.
Working knowledge of agent patterns and their failure modes, and the judgment to know when a deterministic script is the better answer.
Experience with workflow automation (n8n, Zapier, or comparable) and with REST APIs and webhooks.
Hands-on experience with HubSpot or a comparable CRM and marketing-automation platform.
Experience with analytics and tracking (Google Tag Manager, PostHog) and technical SEO tools (Semrush, Ahrefs, Search Console). Familiarity with SQL, BigQuery, and marketing attribution.
Comfortable with modern CMS and hosting stacks (Sanity, Vercel) and translating Figma designs into responsive pages.
Git-based development, staged deployment, and clear communication with technical and nontechnical partners.

Nice To Haves

Experience with evals, observability, or LLM-ops tooling.
Familiarity with MCP or similar tool-integration standards, vector databases, and embedding-based retrieval.
Python for data and AI work.
Exposure to fintech, tax, or another compliance-sensitive domain.

Responsibilities

Build and operate production AI agents that do real marketing work: SEO research, competitive monitoring, content drafting and repurposing, design production, lead enrichment, and inbound triage.
Integrate LLM APIs (Anthropic, OpenAI, and comparable providers) using tool calling, structured output, and retrieval over our brand and product knowledge, orchestrated in n8n, Zapier, or custom services.
Decide where a human stays in the loop, and build the review and rollback paths that make autonomous output safe to publish.
Own reliability and unit economics: monitoring, error handling, API cost tracking, and knowing when a cheaper model or a plain script beats an agent.
Build evals and golden-set test cases for every agent in production, covering accuracy, brand voice, and factual grounding. Treat regressions as bugs.
Version prompts and models so changes are traceable and reversible, and A/B test generated content against human baselines.
Set internal guidelines for AI-generated marketing content, including review of any tax-related claims.
Own technical SEO (metadata, redirects, structured data, sitemaps, Core Web Vitals) and extend it to AI search: how Taxfyle gets retrieved and cited by answer engines.
Build content pipelines that generate, refresh, and internally link pages beyond what a human team could maintain, with quality gates before anything publishes.
Track brand visibility and citation share across AI assistants and traditional search.
Build and maintain marketing sites, landing pages, and conversion flows in TypeScript, React, and Next.js on a headless CMS such as Sanity.
Ship AI-native site experiences where they earn their place (assistive search, guided qualification, personalization), measured on conversion rather than novelty.
Keep performance and reliability healthy, and handle redirects and launches without breaking analytics or search.
Own HubSpot forms, workflows, and lifecycle logic, plus the integrations connecting HubSpot to our product, warehouse, and internal services.
Apply models where rules fall short: lead scoring, intent detection, routing, enrichment, and personalized lifecycle messaging.
Find the root cause of workflow issues before building reusable fixes.
Maintain event tracking (Google Tag Manager, PostHog) and clean marketing data in BigQuery.
Build dashboards on acquisition, funnel performance, and campaign outcomes.
Report AI system performance in the same terms as any other channel: output volume, quality scores, cost per output, and hours returned to the team.