Applied AI Engineer

Arcade•San Francisco, CA

19h•$179,000 - $240,000

About The Position

We are changing the way AI interacts with the real world by creating the infrastructure that will power every AI application. We are building the actions runtime that allows AI agents to safely take real-world actions at enterprise scale. As an Applied AI Engineer on the Tools team, you will push the boundary of what "a tool" means, designing agentic tools that go beyond deterministic API wrappers, building agents that build new tools, and composing tools into workflows that solve higher-level problems. This role is for someone who thrives in ambiguity and wants to ship impactful solutions in a rapidly evolving field.

Requirements

5+ years software engineering experience, with at least 2 years shipping production ML or applied-AI systems.
Strong Python skills.
LLM application depth, including prompting, retrieval, tool use, and agent design. Experience building non-trivial agent systems and understanding their rough edges.
Experience designing or composing multi-tool / multi-agent workflows that produced real outcomes.
Experience building evals at scale, including measurement systems against which real engineering decisions were made.
Statistics fluency, including understanding significance, confidence intervals, and A/B test design.
Comfort across multiple frontier models and reasoning about their behavioral differences.
A do-er mentality, prioritizing shipping working versions quickly over polished long-term releases.
Comfort with ambiguity and making good decisions with incomplete data.
An insatiable desire to ship.

Nice To Haves

Experience building agents that build software (codegen agents, harness-style systems, meta-agents).
Prior work on tool-use specifically (e.g., BFCL, τ-bench, ToolBench, MCP eval work, or equivalent).
MCP ecosystem familiarity, with a bonus for having filed an issue against the spec.
Experience with agent frameworks (LangChain, CrewAI, AutoGen, Mastra) and opinions on their tool use and workflow composition.
Prior experience at an API platform, integrations-heavy product, or developer tools company.

Responsibilities

Design and ship agentic tools that go beyond deterministic API wrappers and define the patterns the rest of the Tools team will use to build more.
Build the agent harness that automates tool creation, taking a vendor's API and producing a high-quality toolkit end-to-end, keeping humans in the loop only where humans add real value.
Design workflows that compose tools into higher-level abstractions customers can actually point at outcomes (e.g., "triage this inbox," "close out this account") rather than individual API calls.
Bring applied-ML rigor to tool design, including evals, model-aware iteration, retrieval, tool description tuning, and response shaping.
Run model-aware experiments across Claude, GPT, Gemini, studying how agentic tool behavior diverges across models.
Set the technical bar for what "good tool-building" looks like as the team scales, ensuring your patterns are inherited by every toolkit author.
Contribute back to the MCP and agent ecosystem where the conversation about agentic tools is forming.