Building — about half your time Integrations between business systems. Data pipelines. Automations that remove manual work from Finance, People Success, IT, Security and GTM. Internal tools and lightweight services. AI-backed workflows and agents where they are the right answer, and conventional code where they are not.
Running it — about a fifth of your time You inherit a set of automations already in production and become their owner. Monitoring, alerting, incident response, and the unglamorous work of making inherited systems observable and documented. Nothing you build gets handed to someone else to support.
Technical evaluation — about 15% When the team is deciding whether to build or buy, or whether a vendor’s API can actually do what the demo claimed, you build the prototype that answers it. Deep technical spikes, integration feasibility, and performance and security reality-checks — against a requirement the Product Manager and Analyst have already defined.
Migrations — about 15% Data migration, field mapping, cutover scripts and system replacements as we move platforms.
The shape of a typical week
Daily — Writing Python with Claude Code. Reading vendor API documentation. Wiring REST and GraphQL APIs, webhooks and event-driven flows. Debugging a sync that silently dropped records. Deploying to Cloud Run.
Weekly — Shipping an increment and demoing it. Working through requirements with the Product Manager and Analyst. Triaging what broke.
Monthly — One meaningful capability live in production, with monitoring, a runbook, and documentation good enough that someone else could pick it up.