Founding AI Engineer — Studio

About Pareto Agent Most AI systems generate text. We’re building one that makes decisions . Pareto Agent is a policy-driven runtime that executes high-stakes commercial workflows—where every action either protects revenue or gives it away. The future of B2B sales isn’t more reps -it is smarter systems . We’re building the execution layer that replaces manual, inconsistent decision-making with deterministic, system-driven outcomes . Inspired by the Pareto Principle (80/20), we focus on the small number of decisions that drive the majority of outcomes—and build systems that execute them with precision. Headquartered in San Francisco, the company is founded by serial entrepreneurs who have successfully scaled multiple B2B companies and has secured over $3.5M in funding from their previous investors. The Role As our Founding Software Engineer — Studio , you will own the full product surface through which customers observe, configure, and trust their autonomous AI sales representative — from the analytics data layer that powers it to the interface that makes it legible. Your mandate is to build the interface that makes an autonomous AI sales representative something operators can genuinely trust with real deals. Reporting directly to the CTO/Co-Founder, you will set the frontend architecture, design the analytics data layer, and establish the patterns and conventions that will outlast your first year. Revenue operators use your dashboards and configuration surfaces to steer a system negotiating real contracts — the clarity and accuracy of what you build determines how well they can do that. What You'll Do Architect and ship the Studio frontend — analytics, configuration, and escalation review — from working prototype to production-quality product Design and maintain the analytics data layer — the materialized views and pre-aggregations that power Studio's dashboards, consuming upstream platform events and optimizing for Studio's query patterns Own the data access layer between the database and the UI: fast, correct, customer-isolated reads that hold up under real account volume Build the configuration UX — a guided selection model that lets customers fine-tune agent behavior without needing to understand the system underneath Collaborate with the founder on UX direction , then make independent implementation decisions without needing pixel-perfect specs Instrument and improve query performance and data freshness as customers and active accounts scale Engage with early customers to validate assumptions and translate real-world usage into product direction Who You Are You've built data-dense B2B interfaces and know the difference between a dashboard that impresses in a demo and one that earns trust in daily use You ask "what decision does this help the user make?" before you ask "how do I build it?" Strong React/TypeScript — you've shipped production frontends, not just prototypes Comfortable with the full stack from PostgreSQL read models to UI components — you own the data layer you consume Experience with data-heavy UIs : filterable tables, drill-down hierarchies, time-series charts, complex form systems Familiarity with row-level security, multi-tenant data access , and PostgREST or equivalent query patterns Familiarity with event-driven read model patterns — CQRS, projections, materialized views — you understand how to build a query layer that's derived, not source-of-truth The taste to know when a prototype is good enough and when it will cost you credibility with an enterprise buyer You're highly self-directed and comfortable operating in ambiguous, early-stage environments You're pushing toward agent-first development — you treat coding agents as an execution layer and invest in the scaffolding, feedback loops, and environment design that let them do reliable, high-throughput work Nice to Have: Domain experience in B2B SaaS, sales technology, or revenue operations — understanding what a revenue operator actually needs from a