Staff Product Architect

About Cadre AICadre AI is an AI strategy and integration firm that builds production AI systems for B2B companies in private equity, wholesale lending, real estate, and SaaS. You will own the technical layer beneath that library. That call sets the hours, and the hours set the margin.You will start as an individual contributor inside Product and Research, then work directly with the head of product engineering and the engineering team directly to build out these reusable components. What You'll DoComponent Taxonomy and Module ArchitectureDecompose every block in Cadre's library into named technical components + product requirements, each with a defined configuration surface, an owner, and a maturity rating.Define the classification framework that determines what qualifies as a reusable module versus a client-specific build.Establish naming, versioning, and documentation standards so any engineer can find, make updates to, or deploy an existing component without asking anyone.Maintain the component catalog as the single source of truth for what Cadre can assemble versus what it must build from scratch.Reuse Economics and PrioritizationTrack attachment rate for every component: how many sold engagements include it, and what that reuse is worth in engineering hours.Prioritize which components get built configurably next, weighing attachment rate, build cost, and margin impact.Report net-new engineering hours as a percentage of total quoted hours, by block and by quarter, as the primary measure of portfolio leverage.Identify blocks that are configuration variants of one another and consolidate them into a single parameterized component.Pre-Sale Solution AssessmentHold final sign-off on how every new client request is scoped: an existing block, an assembly of existing components, or net-new engineering.Act as the ultimate responsible party for guiding the Solution Architecture and engineering teams on when a new technical component should be built, and work closely with Engineering to architect specifically how to build those components.Partner with Strategy and Sales during client deep dives to translate requirements into component-level scope before pricing is set, attending client meetings as needed for direct feedback.Approve engineering hours on pricing sheets so Cadre never quotes a build for infrastructure that already exists in the catalog.Push back when a request is being scoped as bespoke work that existing components already cover.Configurable Build StandardsDefine what parameterized means at Cadre: the bar a component must clear before it counts as reusable rather than forked.Drive high-attachment components from bespoke implementations to productized, configurable modules alongside the engineering organization.Identify which existing builds are strong candidates to be turned into components, but also map new builds against the catalog to identify opportunities for reuse at the outset of building.Own the process for versioning and updating components as client deployments surface new configuration requirements.Recommend upgrades to existing deployments when components are enhanced; partner with Service Expansion team to do so.Establishing the Product Architecture FlowOperate hands-on, reading code and performing the decomposition work directly.Partner tightly with the engineering team to rapidly assess the current state of configurability and establish a clear future state vision.Raise the reuse instinct across Cadre's engineering organization so solutions are built from component configurability.Over time, work with a small team of engineers to build out the highest-leverage components, setting technical direction and reviewing standards as needed.What You Need to Succeed10+ years building production software, including at least 3 in a role where you owned architecture decisions across multiple products or client engagements rather than a single codebase.Proven experience turning bespoke, one-off builds into reusable configurable components, plus the judgment to know which ones were worth productizing and which were not.Deep fluency with modern AI application patterns including LLM orchestration, retrieval, document processing pipelines, voice agents, and agentic workflows, with production systems you can point to.Comfort working directly in the codebase. You will read repos, evaluate abstraction boundaries, and make calls that engineers live with, so technical credibility is non-negotiable.Strong command of the stack Cadre builds on, including TypeScript and Node, Python, Next.js, Postgres or Supabase, and cloud deployment across Vercel, Render, or AWS.Ability to translate technical architecture into commercial terms (hours, cost, scope, and margin) and defend an estimate to a sales team and a client in the same conversation.A track record of establishing standards that people actually follow, rather than documentation that gets written once and ignored.Willingness to hold the line.