Build and scale an AI-powered audit workbench that transforms raw evidence into reviewable workpapers. Responsibilities include designing ingestion and extraction pipelines, deterministic recomputation and exception detection, workpaper generation, evaluation infrastructure, and frontend review tools. The role emphasizes production LLM reliability, golden-dataset evaluations, schema design, traceability, deterministic regulatory logic, structured human review, and end-to-end system ownership in a regulated domain.

Build and own Pepper’s regulated audit platform across internal auditor tools, client and partner portals, data models, integrations, security, and reliability. Develop end-to-end features using TypeScript, React, Node.js, PostgreSQL, and Supabase. Design secure multi-tenant data isolation, authentication, audit logging, evidence integrity, integrations, observability, CI/CD, and deployment workflows. Partner with founders, auditors, CPAs, engineers, and external partners in a high-autonomy environment.