Staff Software Engineer

The role : -

• Youll own engineering work end-to-end across our products and platformfrom ambiguous us problem to shipped feature.
• Translate fuzzy product and engineering asks into shipped systems; you dont wait for a manager to break the work down for you.
• Build with Claude Code, Codex, and the broader agentic stack as your default workflow not novelty.
• Design and ship backend systems, APIs, and data platforms that hold up at scale.
• Partner directly with product owners and stakeholders. Product instinct matters as much as code.

How we engineer : -

• We write evals, not just tests. We engineer the harness: model routing, context and prompt design, tokenomicsfirst-classcraft,not after thoughts.
• Weauth or skills and MCP plug insto specialize our toolchain we dont just consume it.
• Open-weights models get explored when they unlock cost, sovereignty, or capability that frontier APIs cant.
• We treat the engineering loop itself as something to instrument and improve, week over week.

How we operate : -

• Closer to a Forward-Deployed Engineer than a heads-down coder. Twice-daily written updates are the default, not an ask.
• Decisions, trade-offs,and approaches live where every one can see them silence stays comfortable because the writing is already flowing.
• We respond fast on chat and email, and treat documents as how we think and align.

What we look for

• 7-10 years building great software. Stack is secondary; strong fundamentals and learning agility are not.
• Youve already rewired how you work in the last 6-12 months around AI tooling. You can tell us specifically what changed, and why.
• High agency. You spot the problem, frameit, and ship the fix with out waiting to be asked.
• Communicates in writing by default. Surfaces progress, decisions, and blockers fast - not weekly, not when asked.
• No cognitive surrender. You think with AI, not through ityou spot slop, push back on the model, and ship higher-quality work because of the loop, not despite it.
• Sharp judgment on trade-offs, prioritization, and when to push back.
• Comfort with ambiguity and a bias to action.
• Python, Fast API, LangGraph, and Postgres are common in our stackbut wed rather hire a generalist who thinks AI-native than a specialist whod oesnt.