Notch builds AI agents for customer support in compliance-heavy sectors, combining orchestration logic with structured data capture. The tech stack—React + Node.js + PostgreSQL + Redis + AWS container services—reflects a full-stack SaaS architecture built for scale, while active projects around LLM/vector architecture, distributed queues, and AI agent deployment signal heavy investment in reliability and multi-tenant orchestration. Pain points centered on scaling to 100k+ messages/day and complex decision logic suggest the core challenge is making AI agent outputs predictable and auditable in regulated environments.
Notch is a New York-based AI startup (founded 2021, 11–50 employees) that builds customer support agents for enterprises in complex regulatory domains. The product combines AI-driven conversation handling with structured guardrails, compliance oversight, and data standardization—turning support interactions into queryable datasets for downstream analytics and decision-making. The engineering-skewed hiring profile (8 of 13 open roles) with seniority weighted toward senior and lead levels indicates an organization scaling infrastructure and AI capabilities simultaneously, with ongoing expansion into Israel.
Frontend: React, TypeScript, Material-UI, styled-components; Backend: Node.js, Express.js, PostgreSQL, Drizzle ORM, Redis; Infrastructure: AWS (EKS, ECS, Fargate); Observability: OpenTelemetry; Testing: Playwright.
AI agent management platform with LLM/vector-based architecture, multiplayer web architecture, distributed systems (queues, caches, locks), design systems, and end-to-end AI platform infrastructure for handling high-volume customer interactions.
Other companies in the same industry, closest in size