About this role
TypeSafe AI seeks a Member of Technical Staff, Model Capabilities to improve the intelligence and production reliability of its AI models through datasets, evaluation tools and user interfaces. The role combines engineering, data science and product judgment to make model behavior easier to inspect and improve.
Responsibilities
- Create datasets, products and interfaces that enable new model capabilities and use cases. Build internal tools for evaluation, debugging and data inspection.
- Develop and own frameworks that measure model quality, reliability and emergent characteristics across iterations. Analyze and experiment with how data, training choices and targeted interventions affect model behavior.
- Turn raw model outputs and data into clear, inspectable interfaces, owning work end to end. Work with product managers and customers to translate real-world needs into model and system improvements, and iterate on techniques for trustworthy AI products.
Qualifications and tools:
- The team seeks generalists with approximately 2–7 years of experience, strong coding fundamentals, product intuition, attention to detail and the ability to work creatively in ambiguous situations. Hands-on experience implementing LLMs and an understanding of their capabilities and limitations are expected; candidates should examine actual model generations rather than trust LLM outputs blindly.
- The stack is primarily Python, with TypeScript, Next.js and Tailwind CSS for frontend work and Kubernetes for orchestration. Developers may use helpful tools, including Claude Code and Cursor.
Work and compensation:
- Full-time, fully in-person at the San Francisco office near Embarcadero station. Base salary is $150,000–$250,000 plus equity, based on leveling. Benefits include fully covered health insurance, daily lunch and dinner, visa sponsorship and a 401(k) plan.
Skills for this role
PythonTypeScriptNext.jsTailwind CSSKubernetesClaude CodeCursorLarge language modelsDataset developmentModel evaluationEvaluation frameworksData analysisExperimentationDebuggingUser interface developmentData scienceProduct intuition