Founding Machine Learning Engineer

Arcade is seeking a Founding Principal Machine Learning Engineer to build and productionize fast, on-prem agent tool recommendation and runtime ML models for enterprise environments.

Responsibilities

• Own the ML pipeline end to end, from data ingestion and training through evaluation and model release
• Make model shipping routine by improving pipeline reliability and repeatability
• Fine-tune and train models for tool selection, routing, retrieval, and agent memory
• Apply practical modeling techniques such as distillation, embeddings, rerankers, and other approaches as needed
• Extend the system to new domains starting with search and recommendation over tools and agent context
• Build an evaluation system that includes:
  • offline evals against real agent traces
  • online measurement in production
  • head-to-head comparisons versus Claude, GPT, and Gemini
• Create a feedback loop from production agent traces and tool-call data to improve models within enterprise customer data boundaries
• Quantize, optimize, and package models for customer VPCs and air-gapped environments, and collaborate with the Runtime team on serving
• Define the ML stack and strategy, including helping decide who to hire next into ML
• Use expertise in modern agent systems (harnesses, memory, skills) to identify where models materially improve agent capabilities and to call early whether approaches will work
• Accelerate iteration so projects that take a week today take a day next time
• Operate the model pipeline lifecycle for models across embedding, reranking, classification, recognition, and additional tasks
• Stand behind model quality using the provided data coverage, including reviewing results in detail (for example, row-by-row in a spreadsheet)
• Ensure everything runs where customers run: their cloud, their hardware, and sometimes air-gapped networks
• Lead development of agent capabilities beyond API calls, focusing on intelligent and efficient abilities that outperform competitors
• Own the models that sit in the runtime path of every agent call served
• Productionize the first agent recommendation model, decide the Arcade ML stack, and set forward-looking patterns
• Own build-buy decisions for the stack with an appropriate budget for execution
• Report directly to the Head of Engineering

Requirements

• 7+ years of software engineering experience
• 4+ years training and shipping production ML systems
• Trained or fine-tuned models that reached production and moved a customer metric, not only leaderboard results
• Production experience with agent systems: harnesses, memory, skills, tool use, and sub-agents
• Knowledge of when and how fine-tuning works, including practical application
• Ability to defend evals in review using statistics fluency to judge whether deltas are real or noise
• Experience deploying under real constraints such as tight latency, limited GPUs, or infrastructure constraints (for example vLLM, ONNX, TensorRT, llama.cpp, or equivalents)
• Python for training and ML work, plus TypeScript or Go for production services that serve models
• Capacity to choose the stack, document it, and defend the decision over time
• Doer mindset: prefer shipping working v0.5 quickly over polishing later
• Comfort with ambiguity in an early team setting where the charter expands and decisions are made with incomplete data
• Strong motivation to ship

Technologies

• Python, TypeScript, Go
• vLLM, ONNX, TensorRT, llama.cpp
• Claude, GPT, Gemini
• MCP

Compensation and Benefits

• San Francisco, CA (onsite)
• USD 230,000 per year base salary starting point
• Equity and competitive benefits
• In-person at the San Francisco office

Traction

• Real deployments with Fortune-100 customers, including Morgan Stanley and Open Table

Founder-Market Fit

• CEO previously founded Stormpath (acquired by Okta), creating the first Authentication API for developers
• CTO previously led the vector database team at Redis, shipped 100+ LLM applications, and contributes to LangChain and LlamaIndex

Dream Team

• Engineers from Okta, Redis, Microsoft, Splunk, Ngrok, Google, Airbyte, Disney, and HPE with experience building and founding multiple developer platforms

Bonus Points

• Shipped ML into enterprise on-prem or regulated environments (financial services, healthcare, government)
• Tool-use benchmark or eval experience (BFCL, τ-bench, ToolBench, MCP evals, or equivalent)
• Familiarity with the MCP (Model Context Protocol) ecosystem, including the possibility of filing issues against the spec
• Built training-data pipelines from production traces with privacy controls
• First ML hire experience at an earlier-stage company, with willingness to do things differently
• Open-source contributions or published work that held up with other engineers
• Experience at an early-stage startup and enjoyment of that environment

Application context: Arcade is hiring a principal machine learning engineer to build the future of agent capabilities, with a first agent recommendation model already built (patent pending) and a training pipeline needing productionization. Compensation is aligned to background, experience, and performance within the stated range.