דרושים » הנדסה » Senior Infrastructure Developer

Were looking for an AI Builder to serve as the technical backbone of Nomas internal AI function. Youll build the LLM-powered agents, automation pipelines, integrations, and infrastructure that teams across the company rely on every day.

This is a rare opportunity to build a new function from the ground up within a company where the work can have an immediate and meaningful impact. Youll create systems such as tools that help Sales track competitive deals in real time, pipelines that automatically turn customer calls into Salesforce updates, and agents that identify competitor activity and deliver actionable insights directly to Slack.

This is a highly hands-on role with true end-to-end ownership-from translating business needs into focused technical solutions to deploying, monitoring, and continuously improving production systems.

What Youll Do:

Build AI-Powered Agents and Automations
Design and build LLM-powered agents, automation pipelines, and backend services used across the organization.
Develop complex integrations that bring together data from multiple internal and external systems.
Build custom MCP servers and other infrastructure required to support scalable internal AI workflows.
Turn fuzzy business needs into practical, focused, and maintainable technical solutions.

Own the Internal AI Infrastructure
Harden and maintain Nomas existing MCP Gateway ecosystem.
Improve reliability, error handling, authentication, secrets management, and versioning.
Implement secure, vault-based credential management and strong security practices across production automations.
Build infrastructure that enables agents and workflows to operate reliably at scale.

Drive Production Readiness and Reliability
Own the deployment, monitoring, maintenance, and reliability of agents and automations in production.
Build evaluation frameworks, test sets, and monitoring processes to measure accuracy, precision, recall, hallucination rates, and overall agent performance.
Identify regressions and continuously improve the quality and stability of production workflows.
Troubleshoot failures across integrations, mod

is looking for a DevOps engineer to own our infrastructure end to end.
You would be our only one. There is no platform team to join and no one to hand you a ticket queue. There is plenty of work, and most of it is on this page. You will set the technical direction as we grow, report to the Head of Engineering, and work directly with every engineer on the team.
We run on AWS and Kubernetes and deploy through GitOps. Our use of AI in engineering is growing quickly, and the infrastructure behind it has not caught up yet. Building that is a real part of this job.
What Youll Do:
Own all of our infrastructure. Kubernetes, AWS, and the core services around them. The foundation was built properly and the team has kept it running between other priorities. You are not inheriting a mess. What it has not had is a full-time owner. Reliability, cost, security posture, and the roadmap all become yours.
Own CI/CD. Every services build and deploy pipeline, and the self-hosted runner infrastructure underneath. We ship fast and we are working on shipping faster. One of our engineers is automating more of it right now, and you will take that over and push it further. The interesting problem is adding the safety a growing company needs without turning deploys into something people dread.
Help us shape the infrastructure behind our AI development. Our engineers use coding agents every day and that use keeps growing. We do not have a clear path for the infrastructure behind it yet, and you will help us find one. One direction we are exploring is moving agents off laptops and into the cloud, with a sandbox environment per engineer so an agent keeps working after the laptop is closed. Deciding whether that is the right answer, and building what we choose, is part of this job.
Scale how we deploy to customers. Some customers need us to run inside their own environment. We already do this, but each deployment takes too much hands-on work. You will make it repeatable, faster, and secure, and give our engineers the tooling to support customers directly.
Own our observability. Metrics, logs, alerting, and dashboards. Decide what is worth waking someone for, and keep the cost of watching under control.
Own infrastructure security and support our compliance work. Hardening, IAM, secrets, and vulnerability management. You will help us keep our SOC 2 compliance, which customers require before they will work with us, and help answer their security reviews. We sell to security teams, so they ask hard questions.
Keep our data recoverable. Backups, restore testing, and a disaster recovery plan for RDS Postgres and ClickHouse that we have actually rehearsed.
Partner with engineers on solutions, not tickets. You are the person developers come to when they need infrastructure to do something new.