Founding Product Engineer (Full Stack) -- AI Platform

## About the Role WindShift is building the Operating System for AI-Native Investment Diligence. We are redesigning execution around AI as a core primitive — not layering it onto legacy workflows. The platform is live with paying customers. The architecture exists. We are scaling the AI-native execution layer into a durable enterprise system. We are hiring a founding engineer who builds systems the AI-first way. This role will help define the execution architecture that underpins a venture-scale AI-native platform. ## What AI-First Means Here We do not want manual feature builders. We want engineers who: - Design systems where LLMs are orchestration primitives - Structure execution as agent graphs, toolchains, and evaluators - Treat Claude/Codex as collaborators, not autocomplete - Optimize for system leverage, not lines of code - Build evaluation loops, regression harnesses, and hallucination detection - Think in terms of compounding structured intelligence If your default mode is writing everything manually before asking whether AI should orchestrate it, this role is not a fit. ## What You'll Architect ### AI-Native Execution Layer - Multi-step agent pipelines (planning → retrieval → synthesis → verification) - Tool and adapter abstractions that allow structured orchestration - Canonical intermediate representations decoupled from rendering ### AI-Driven Development Workflow - Structured prompting systems - Internal orchestration layers around Claude/Codex - Automated evaluation and regression testing - Cost-aware model routing and caching strategies ### Production-Grade Infrastructure - Async worker reliability (retries, idempotency, backpressure) - Observability and traceability across agent runs - Multi-store data architecture (Postgres, Redis, vector DB, graph DB) - Terraform-managed AWS deployment (ECS, RDS, ALB) ## The Standard You should: - Have 5+ years shipping production systems - Be fluent in modern LLM tooling and orchestration frameworks - Understand distributed systems tradeoffs - Be comfortable designing architecture, not just implementing tickets - Operate independently in a high-context environment You should not need step-by-step specs. ## Why This Is a Rare Opportunity AI-native execution will reshape professional services economics. WindShift already compresses repeatable workstreams by 20–30% in live environments. The next phase is scaling that leverage structurally to >50%. With a small, high-talent team, AI-first engineering is the multiplier. This role is foundational to that leverage. ## Ownership - Direct architectural influence - High autonomy - Meaningful equity - Path to long-term technical leadership We are building a durable AI-native company, not an AI feature layer.