Software Engineer - AI Tooling

About this role

Build and improve AI tooling that helps developers and agents work with Supabase, including agents, an MCP server, agent skills, and a dashboard assistant. The role emphasizes evaluation, instrumentation, reliability, and enabling new Supabase products and customer applications to work well with AI agents.

About the company

Supabase builds an open-source backend platform with Postgres, auth, storage, and realtime APIs. It helps developers launch production apps without managing custom infrastructure.

The team

AI Tooling team

Stack

What you'll do

• Build and maintain the MCP server, agent skills, agents, and dashboard assistant.
• Create tools, skills, and evaluations for new Supabase product surfaces and emerging AI technologies.
• Develop evaluations, metrics, instrumentation, and quality gates to measure AI task success and prevent regressions.
• Use customer and agent feedback to improve tooling, interfaces, documentation, and product knowledge access.
• Build patterns for AI features in customer applications, including MCP, embeddings, and agents on Edge Functions.
• Create contribution paths and shared AI logic that help teams improve and reuse AI capabilities.
• Test, benchmark, and stress-test tools; collaborate on production reliability, authentication, and hosting.

What we're looking for

• 5+ years of professional software engineering experience.
• Strong TypeScript skills and familiarity with the JavaScript ecosystem, including npm, ES modules, and bundlers.
• Deep familiarity with the AI ecosystem, agent patterns, and practical trade-offs.
• Experience building developer tools, product interfaces, or platform primitives for engineers.
• Experience designing or using AI evaluation frameworks; understand baselines, variance, cost, and task success.
• Strong testing and reliability practices, including real-world validation.
• Knowledge of Postgres and relational database practices.
• Clear written communication and ability to work across teams in an async, distributed environment.

Nice to have

• Experience with vector databases and embeddings, especially pgvector.
• Examples of shipped customer-facing AI features or agents.
• Experience designing developer or agent-focused documentation and structured knowledge.
• Experience with Braintrust or similar evaluation and tracing platforms, or CI quality gates.
• Open-source contributions or maintained developer tooling.
• Experience working across multiple product surfaces or teams.

Benefits

• Company ESOP equity.
• WeWork membership or coworking allowance.
• Tech allowance for work equipment.
• Health insurance coverage for employees and partial coverage for dependents.
• Annual company off-sites.
• Flexible, asynchronous work.
• Annual professional development allowance.