Member of Technical Staff (Model Capabilities)

Hone; San Francisco (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Hone is hiring a Member of Technical Staff for its stealth frontier model lab. You’ll ship data-tooling interfaces and progress model intelligence through datasets, evaluation tooling, and interfaces, primarily using Python with TypeScript/Next.js on frontend and Kubernetes for orchestration.

You’ll build high-leverage datasets, evaluation frameworks, and UIs, partnering with product teams and customers to translate real-world needs into production-ready reliability for autonomous workflows.

Qualifications

• Generalist with solid fundamentals and 2–7 years of experience.
• Strong coding ability and product intuition.
• Willingness to ship data-tooling interfaces quickly and iteratively.

Responsibilities

• Create high-leverage datasets, products, and UIs on top of the model.
• Build internal data tooling and evaluation interfaces for model development.
• Develop and own evaluation frameworks to measure quality and reliability across iterations.
• Run analyses to understand data, training choices, and interventions on model behavior.
• Turn raw model outputs into inspectable UIs with domainend-to-end ownership.
• Collaborate with product managers and customers to translate needs into model improvements.
• Continuously iterate to improve trustworthy AI products.

Skills

Generalist

Coding ability

Product intuition

Attention to detail

Tools

Python

TypeScript

Next.js

Tailwind CSS

Kubernetes

Job description

A stealth frontier model lab is hiring a Member of Technical Staff. The company is pre-launch, sub-15 people, backed by top-tier investors, and heading toward a major general release in fall 2026. The founding team comes from OpenAI, Google Brain, and Meta FAIR. They're solving one of the hardest problems in AI: making LLMs reliable enough to build real autonomous workflows on.

About the Role

They're looking for scrappy engineers who ship data-tooling interfaces fast. You will progress model intelligence through the interfaces, datasets, and evaluation tooling you build, combining product sense, data science, and engineering. You'll be part of the real "secret sauce" of the lab: figuring out how their models can provide production-ready reliability in the real world.

Their tech stack is primarily Python. They also use TypeScript, Next.js, and Tailwind CSS for frontend, with Kubernetes for orchestration. They empower engineers to use any tooling they find helpful for getting their job done, including Claude Code and Cursor.

What You'll Do

• Create high-leverage datasets, products, and user interfaces that unlock new capabilities and use cases on top of their model
• Build internal data-tooling and evaluation interfaces quickly, making model development clearer (evaluation, debugging, data inspection)
• Develop and own evaluation frameworks to measure quality, reliability, and emergent characteristics across model iterations
• Run rigorous analyses and experiments to understand how data, training choices, and targeted interventions impact model behavior
• Turn raw model outputs and data into clear, inspectable UIs, owning a domain end-to-end, with craft and judgment over simple optimization
• Partner closely with product managers and customers to translate real-world needs into concrete model and system improvements
• Continuously iterate, learn, and co-discover new techniques for building exceptional, trustworthy AI products

Who You Are

• A generalist (~2 to 7 yrs) with solid fundamentals, strong coding ability, and product intuition
• Maintain high attention to detail and a strong bar for quality
• Enjoy thinking beyond the code to how systems are used in the real world
• Scrappy and hacky, and thrive in ambiguous problem spaces, taking satisfaction in finding creative solutions
• Don't trust LLMs blindly: you look at the reality of model generations
• Have hands-on experience implementing LLMs and understand their capabilities and limitations