Software Engineer - AI Tooling

This is a fully remote position, open to applicants anywhere in the world.

📋 Description

• Develop and enhance AI tools, including the MCP server, agent skills, agents, and the Assistant dashboard.

• Create new Supabase products that agents can utilize through tools, skills, and assessments.

• Prototype and assess tools for emerging AI technologies to identify candidates for production deployment.

• Construct evaluations, instrumentation, feedback loops, quality KPIs, and regression gates.

• Examine customer and agent interactions with deployed tools to enhance tools, interfaces, and documentation.

• Design the presentation of documentation and product knowledge to agents while measuring effectiveness.

• Develop patterns and tools such as MCP on Edge Functions, vector embeddings, and agents operating on Edge Functions.

• Facilitate contributions to Supabase’s agentic capabilities via contribution paths, templates, documentation, and review processes.

• Centralize shared AI logic for reuse across various teams.

• Deliver robust tools that have been rigorously tested, including thorough benchmarking and stress testing.

• Collaborate with AI Tooling Infrastructure engineers on aspects of reliability, authentication, and hosting.

• Work alongside Docs, CLI, Frontend, SDK, and other teams to offer a cohesive developer experience.

• Keep abreast of the rapidly evolving AI landscape.

⛳️ Requirements

• Over 5 years of professional software engineering experience.

• Proficient in TypeScript with a solid understanding of the JavaScript ecosystem, including npm, ES modules, bundlers, and compatibility.

• Extensive knowledge of the AI ecosystem, current tools, agent patterns, and practical trade-offs.

• Experience in building developer tools, product interfaces, or platform primitives utilized by engineers.

• Familiarity with designing or employing evaluation frameworks for AI quality, reliability, and task success.

• Understanding of baselines, variance, and costs associated with AI evaluation.

• Strong judgment regarding effective AI workflows to avoid fragile or gimmicky experiences.

• Proven experience in writing tests and validating reliability under real-world conditions.

• Knowledge of Postgres and best practices in relational databases.

• Excellent written communication skills and the ability to collaborate across product, engineering, and documentation teams.

• Ability to excel in asynchronous, globally distributed teams.

• Comfort in navigating uncertainty and acting swiftly.

• Nice to have: experience with vector databases and embeddings, particularly pgvector.

• Nice to have: experience in delivering customer-facing AI features or agents.

• Nice to have: documentation or structured knowledge designed for developers or agents.

• Nice to have: experience with Braintrust or similar evaluation and tracing platforms, or CI quality gates.

• Nice to have: contributions to open-source projects or maintained developer tools.

• Nice to have: experience across multiple product surfaces or teams.

🏝️ Benefits

• Fully remote work environment.

• WeWork membership or co-working allowance applicable worldwide.

• Employee Stock Ownership Plan (ESOP).

• Technology allowance for a laptop, monitor, headphones, or other work-related equipment.

• Health insurance coverage: fully covered for employees and 80% for dependents.

• Annual company off-site events.

• Flexible, asynchronous working hours.

• Annual education allowance for courses, books, conferences, or other learning opportunities.