• Plan, conduct, and coordinate Agile software development activities, focusing on the maintenance and sustainment of NCEI’s enterprise archive systems • Apply the appropriate standards, processes, procedures, and tools throughout the software development lifecycle • Manage the lifecycle of change requests, deploying changes across various environments • Ensure software standards are met • Correct program errors, prepare operating instructions, compile documentation of program development, and analyze system capabilities to resolve questions of program intent, output requirements, input data acquisition, programming techniques, and controls

• Own outcomes end-to-end, wherever you're deployed. • Frame the problem, design it, ship it, own the metric. • Descope to the cheapest version that works. • Prototype in days. • Stay close to customers. • Shadow sales and support, join advisory conversations, and let customer reality drive technical decisions. • Propose what to build, defer, and kill. • Own your domain and share it. Its interfaces, reliability, performance, and security, including on-call for the systems you own. • Consent and PII handling are central to customer trust. • Bring patterns, tooling, and hard-won context back to the wider org rather than hoarding them. • Build and scale both the product surface and the platform beneath it. • Ship responsive, accessible React interfaces with a strong UI/UX eye, back them with Node.js and Python services and APIs, and keep the cloud batch platform and data pipelines that deliver multi-tenant products reliable as they scale. • Architect AI-driven leverage. • Build AI-assisted workflows that hold up, evolve the team's practices as tooling changes (Cursor, Claude Code, and whatever comes next), and look for product wins AI unlocks (smarter routing, anomaly detection, and more). • Multiply the teams you touch. • Mentor on judgment and product sense, not just syntax. • Write decision memos and ADRs. • Enable Product and Design to self-serve.

• Own the design, build, and operation of Kubernetes cluster management tooling and automation that keeps our compute platform reliable and self-healing at scale. • Build developer-facing tooling and workflows that improve how engineers across Coinbase interact with Kubernetes, with a heavy emphasis on integrating AI-driven processes and support. • Deliver net-new compute capabilities for service owners, such as one-off jobs, cron scheduling, deployment strategies, EFS support, and automated right-sizing. • Drive operational excellence by automating toil, reducing on-call burden, and continuously improving platform observability and incident response. • Partner with Security, Reliability, and Observability teams to ensure the compute platform meets Coinbase's standards for security, uptime, and performance.

• Design, build, and ship product features across the Fireworks platform — including the developer console, API surfaces, model playground, fine-tuning workflows, deployment management, and billing • Own features end-to-end: scope the problem, build the solution, ship it, instrument it, and iterate based on real usage • Build and improve the developer experience for Fireworks' core product capabilities — serverless and on-demand inference, supervised and reinforcement fine-tuning, multi-LoRA serving, and agentic orchestration • Work across the full stack — frontend (React/TypeScript), backend services (Python, Go), APIs, and data layers — to deliver polished, production-quality product surfaces • Build for products that are evolving rapidly: new model types, new serving paths (standard, priority, fast), new consumption models like Fire Pass, and new capabilities like multi-agent harness tooling • Partner closely with Infrastructure, Applied Research, Growth, Data and Design to translate platform capabilities into intuitive developer-facing products • Instrument and monitor the systems you build, using data to understand how developers use the platform and where friction exists • Move fast and make pragmatic trade-offs — the pace of AI infrastructure demands shipping early, learning quickly, and iterating with conviction