Founding AI Engineer - Paradox Machines

Infinity•New York, NY

2d•Remote

About The Position

This is a founding engineering role at a data and AI company - which means you'll be close to the product, close to customers, and building systems that are used from day one. You'll design and ship the AI-powered applications and data infrastructure that sit at the core of what we sell: automated reporting, intelligent data pipelines, and AI layers on top of operational systems. You're not just writing features - you're helping build the platform Paradox runs on. You'll work directly with the CEO and alongside a small, senior team. The problems are real, the feedback loop is short, and the surface area is large. Expect to move fast and own a lot. What You'll Build The data platform You'll design and maintain the infrastructure that powers client deliverables - ingestion pipelines, data models, transformation layers, and the orchestration that keeps it all running. You understand how data moves, where it breaks, and how to build systems that hold up under real conditions. dbt, Airflow, Spark, cloud data warehouses - you’ve either worked with, or understand conceptually, how these systems integrate and when to use what. AI-native applications You'll build and ship AI applications - not prototype them. That means writing production-grade code, designing agent workflows, and integrating LLMs into systems where accuracy and reliability actually matter. You understand the difference between a demo and something that works at scale, and you build for the latter. Agent systems that stay accountable You believe in moving fast, but not at the cost of reliability. You'll architect agent systems with clear evaluation criteria, human-in-the-loop checkpoints where they matter, and monitoring that tells you when something goes wrong. AI is a multiplier - you use it as one. Production infrastructure You'll own deployment end-to-end - containerized services, Kubernetes orchestration, cloud infrastructure, and the CI/CD pipelines that keep everything moving. You think about reliability, scalability, and cost from the start, not as an afterthought. Internal tools and reusable assets As a founding engineer, you'll help identify what we build repeatedly across engagements and turn it into durable internal infrastructure - shared data models, evaluation frameworks, deployment patterns. You're building for the long game, not just the current sprint. What You Bring

Requirements

5-8 years of software engineering experience, with meaningful time and familiarity with building data systems or AI-driven applications in production.
Hands-on experience with modern data engineering practices: pipeline orchestration, data modeling, transformation frameworks, cloud infrastructure. You don’t need to be a data engineer, but we are building a data platform.
Experience deploying and operating production systems - Kubernetes, Docker, CI/CD, cloud platforms (AWS or GCP). Pipelines break. We need to fix them.
Experience building with LLMs in production - prompt engineering, RAG, agent frameworks, evaluation pipelines. You've shipped things that real users rely on. You will build both with AI and build the agents within the platform.
Comfort designing agentic workflows where the AI does meaningful work but accountability stays with the system. You are accountable for everything AI does.
A bias for doing: you'll write the query, sketch the schema, or spin up the prototype if that's what moves things forward.
Genuine curiosity about what AI can do next - and the technical discipline to build it responsibly.

Responsibilities

Design and maintain the infrastructure that powers client deliverables - ingestion pipelines, data models, transformation layers, and the orchestration that keeps it all running.
Build and ship AI applications - not prototype them. That means writing production-grade code, designing agent workflows, and integrating LLMs into systems where accuracy and reliability actually matter.
Architect agent systems with clear evaluation criteria, human-in-the-loop checkpoints where they matter, and monitoring that tells you when something goes wrong.
Own deployment end-to-end - containerized services, Kubernetes orchestration, cloud infrastructure, and the CI/CD pipelines that keep everything moving.
Help identify what we build repeatedly across engagements and turn it into durable internal infrastructure - shared data models, evaluation frameworks, deployment patterns.