Job Description
Job Description
Creao is building the next generation of AI-native workspaces, where intelligent agents drive collaboration and creation. We’re moving beyond “apps built by prompts” toward a shared system of configurable, composable skills that teams can evolve together.
As a Machine Learning Engineer (Agent Systems), you’ll design and scale the intelligent runtime that powers these agents — enabling long-running, observable, and adaptive AI workflows. You’ll help shape how humans and AI co-operate inside a workspace designed for speed, reliability, and creativity
Key Responsibilities
-
Design and implement the runtime architecture for AI agents that can execute multi-step, contextual tasks.
-
Develop systems that turn structured skill configurations into working, reusable agent behaviors.
-
Build the infrastructure for long-running and interactive AI tasks, with strong reliability and observability.
-
Integrate and optimize large language models and other ML components for task orchestration and reasoning.
-
Collaborate cross-functionally to ensure smooth interaction between AI systems, data layers, and user interfaces.
-
Drive improvements in performance, scalability, and model efficiency across the agent stack.
What You’ll Work On
-
The agent runtime that coordinates intelligent task execution.
-
Frameworks that make AI skills composable and reusable across teams.
-
Mechanisms for safe, transparent, and human-in-the-loop AI operations.
-
Experiments in adaptive planning, retrieval, and multi-agent collaboration.
You Might Be a Fit If You Have
-
Experience building or scaling machine learning or LLM-based systems.
-
Strong software engineering fundamentals (Python, Typescript, distributed systems, ML infrastructure).
-
Understanding of stateful task execution, streaming inference, or agentic orchestration.
-
Ability to work across the stack — from model integration to API and runtime design.
-
A passion for creating usable, reliable, and intelligent AI systems that empower others to build faster.