What you'll do
- Talk to people, then build things. You'll work directly with business and engineering teams to understand what's slowing them down. You'll figure out whether the fix is traditional software, an AI integration, or some combination. Then you'll turn vague requirements into something you can actually build.
- Own the whole thing. Prototype it, harden it, deploy it, monitor it. These are internal tools, so you move fast, but they still need to work reliably.
- Build fast with AI-assisted development. We use tools like Claude Code to move quickly. You should be comfortable using them to scaffold, iterate, and ship. But you also need to read what they produce, catch what they get wrong, and know when to write it yourself. These tools make good engineers faster. They don't replace knowing how software works.
- Use AI where it actually helps. Integrate LLM APIs (OpenAI, Anthropic, Gemini, etc.) when language understanding or probabilistic reasoning genuinely improves the outcome. Design prompt strategies and evaluation methods when you do. Skip the model call when a conditional statement would work better. Keep an eye on cost, latency, and reliability.
- Write code other people can maintain. Build clean systems. Establish practical patterns for secure AI usage. Contribute to standards around observability, safety, and data handling.
- Work with data. Write SQL against MySQL/ Snowflake, build internal dashboards, and turn business questions into lightweight data tools.
What we're looking for
- 5+ years of professional software engineering. You've shipped production systems used by real people.
- Solid backend skills: APIs, data modeling, system design.
- Hands-on experience integrating LLM APIs into real applications.
- Fluent with AI-assisted dev tools (Claude Code, Codex, Cursor, or similar).
- Strong SQL (Snowflake experience is a plus). Javascript or Python preferred.
- Comfortable with ambiguity. You don't need perfect specs to start building, and you take ownership from idea through production.
- You can tell the difference between interesting and impactful. You talk to stakeholders directly and optimize for usability.
What success looks like after 6 months
You've shipped multiple internal tools that teams actually use on a regular basis. You've identified and fixed real workflow bottlenecks. You've helped set pragmatic standards for how engineering uses AI. People trust you to build things that work.