About the role
We’re looking for a full-stack engineer to work closely with our client on several AI-heavy projects running in parallel. This isn’t a typical feature-shipping role. Much of your time will go into designing and building AI feedback loops: systems where autonomous coding agents run continuously, and your job is to make them productive, reliable, and well-supervised.
You’ll work across:
Computer vision web apps: web applications built around image and video understanding
Insurance automation data pipelines: pipelines that ingest, process, and automate insurance workflows
React Native apps: several mobile apps that will be built largely by an AI agent. Your focus will be building and improving that agent, not hand-writing every screen.
What you’ll do
Design and run AI-assisted development loops, including agents that operate around the clock
Build the tooling, evaluation, and feedback mechanisms that keep those agents on track
Architect and ship AWS-native solutions, including containerized workflows
Move between multiple projects regularly and keep momentum on each
Collaborate closely and communicate clearly in English throughout the day
What we’re looking for
Required
5+ years of professional full-stack development experience
Strong proficiency in at least two of: Python, Ruby, JavaScript
Hands-on experience with AWS-native architectures
Solid understanding of container workflows (Docker, ECS/EKS or similar), and the ability to discuss them in depth
Practical experience with AI-assisted development: working with coding agents, structuring their tasks, and managing long-running or continuous agent loops
Good written and spoken English
Ability to work with substantial overlap with CST business hours (at least 3-4 hours of overlap)
Comfort with frequent context switching across several active projects
Uninterruptible power and internet setup, staying far from the war zone (in case of living in Ukraine)
Nice to have
React Native experience
Background in computer vision or ML-powered applications
Experience building data pipelines, especially in insurance or other regulated domains
