Staff Innovation Engineer, Marketplace

- Platform & primitives: Design and build the durable, reusable services that every Marketplace product depends on - things like purchase orders, contract terms, invoicing rails, and workflow/status state machines - using fine-grained microservices and asynchronous workflow orchestration (e.g., Temporal). This work is judged on quality attributes: maintainability, scalability, reliability, and flexibility.

- New product bets: Partner directly with customers to scope a problem, ship a working (not perfect) version of a new marketplace product within about 30 days, and iterate weekly based on direct customer feedback. Form and test hypotheses with discipline - know when to keep pushing, when to pivot, and when to walk away.

- Adoption & optimization: For products that have already launched, focus on quality, craftsmanship, and driving adoption and usage - optimizing workflows and click paths and making the product the best version of itself for customers already using it.

- AI-native development: Build and refine agent harnesses and layered prompting systems (system, policy, and steering prompts) that automate real workflows - for example, extracting and reasoning over unstructured documents (loan covenants, plans and specs, appraisals) using a mix of AWS Textract, GCP Document AI, and direct calls to models like Claude, Gemini, and OpenAI, choosing the right tool for the document and cost profile.

Build and use internal eval tooling to test and improve these systems before and after they ship.

- Partner with product, design, sales, and the General Manager to understand customer problems and give input on pricing, GTM, and scale/pivot/kill calls for the bets you're closest to - final commercial decisions rest with the GM and PM, not the engineer.

- Mentor other engineers on the team through code review and pairing, and contribute to system design and architecture decisions even on bets you aren't personally driving.

- Maintain the team's service and product-API boundaries (services don't call services; product APIs stay one-to-one with the user journeys they serve) as the team scales its footprint.

- 7+ years of experience in software engineering, with strong backend/distributed-systems fundamentals - designing services and APIs that need to be reliable and maintainable, not just prototypes.

- Hands-on experience building and shipping 0-to-1 products or features in a fast-iteration environment with real customers, ideally using design-partner or similar validation approaches.

- Experience building with or integrating multiple LLM providers (e.g., Claude, Gemini, OpenAI) and document-extraction tooling (e.g., AWS Textract, GCP Document AI), including knowing when to fall back from a structured extraction tool to an LLM.

- Experience building agent harnesses, RAG pipelines, or other AI/LLM-integrated workflows, along with the evals or testing practices needed to trust and improve them.

- Data engineering, ML, or heavy document-processing background is a strong plus - a meaningful share of this team's work is reasoning over unstructured, non-standardized documents (legal terms, covenants, inspection reports).

- Comfort working across the stack as needed - this is a full-stack-capable engineering role, not a narrowly scoped frontend or backend specialization - plus genuine curiosity about the customer problem, low ego, and a preference for working closely with product rather than in a handoff model.

- You've made a substantial, visible contribution in at least one of the team's three workstreams - a platform primitive that other bets now build on, a new product taken from hypothesis to a validated pilot with real customers, or a measurable adoption/usage improvement on a live product.

- You've built or meaningfully improved at least one agent-driven or AI-integrated system that's in production use.

- Engineers and cross-functional partners on the team would describe you as a strong collaborator who raises the bar on system design and is genuinely curious about the customer problem, not just the code.

- If you mentored the team's newer engineer(s), they're visibly applying stronger patterns in their own work.

Built's salary range for this position is $180,000-$240,000 USD per year. The pay range is designed to accommodate upward mobility in the role; therefore, it encompasses the full span of proficiency levels for this role and we believe that the midpoint of the range is competitive in the market. Salary is just one component of Built's total compensation package for employees; your total rewards package at Built will include equity, market-current medical, dental and vision coverage, an unlimited PTO policy, and other benefits.