About the role
About Numeral:
- Numeral is transforming how taxes get done. Digital businesses are currently bogged down by a painful web of regulations across 47 states and 50+ countries. We’re eliminating this burden so teams can focus on their core mission.
- We’re the largest and fastest growing AI-native tax solution. Started in 2023, Numeral has raised over $57M from Benchmark, Mayfield, Y-Combinator, and many others. We now serve over 2,500 paying customers and have more than tripled our revenue every year in our history.
- For this role we’re primarily hiring in our NY office, but do hire remotely in some cases
About the role:
We’re looking for a Foundational AI Engineer who thrives in ambiguity, enjoys building real-world AI systems end-to-end, and wants to define the technical and architectural foundations of an early-stage startup.
This is a hands-on role for someone who wants to design and implement LLM-powered automation systems: model orchestration, tool calling, MCP-style integrations, agentic workflows, and automation-driven tooling.
What you’ll do:
- Design, build, and operate production AI systems, including tool-calling agents, model orchestration layers, and RAG.
- Implement and maintain AI integrations using modern frameworks, MCP servers, and external model providers.
- Build and evolve scalable backend services and APIs that power automated workflows and customer-facing features.
- Think in automation and workflows: design systems using tools like n8n, Claude Code-style agents, or similar orchestration approaches.
- Make architectural decisions that balance latency, cost, reliability, and iteration speed.
- Own infrastructure, observability, and performance for both AI and non-AI services.
- Collaborate closely with product, design, and data teams to ship features that solve real customer problems.
