About the Role
We're a fast-moving startup building a high-traffic communication product and production-grade ML models. We're looking for a language-agnostic DevOps Lead to take complete ownership of our infrastructure from scratch — cloud environments, secure architectures, CI/CD, observability, and developer tooling that backend and ML engineers rely on daily.
Responsibilities
-
Architect, provision, and scale cloud infrastructure across GCP/AWS/Azure for both web backends and compute-intensive ML workloads
-
Establish cloud accounts from scratch with strict IAM policies, RBAC, and secure networking (VPCs, subnets, peering)
-
Build robust, language-agnostic CI/CD pipelines using GitHub Actions, Jenkins, or GitLab CI for safe multi-environment deployments
-
Containerize and orchestrate applications using Docker, Kubernetes, and Helm
-
Deploy and manage Matrix homeserver infrastructure (Synapse/Dendrite) including federation, workers, and database setup
-
Build observability from a blank slate — monitoring, logging, alerting using Prometheus, Grafana, ELK, or Datadog
-
Instrument applications for deep insights into API performance, system health, and ML model inference metrics
-
Manage infrastructure as code with Terraform or Pulumi — version-controlled, repeatable, cloud-agnostic
-
Collaborate with software engineers and ML researchers to remove deployment bottlenecks and maintain developer velocity
Requirements
-
8+ years in DevOps, SRE, or Cloud Infrastructure with holistic infrastructure ownership
-
Language-agnostic — comfortable navigating Python, Go, Node.js, Java codebases for tooling and instrumentation
-
Deep hands-on experience with Kubernetes, Docker, and Helm in production
-
Strong IaC proficiency (Terraform or Pulumi)
-
Extensive CI/CD pipeline experience (GitHub Actions, Jenkins, GitLab CI)
-
Security-first mindset — cloud networking, IAM/RBAC, secrets management
-
Proven ability to build monitoring/logging stacks from scratch
-
Builder's mentality — comfortable starting with a blank canvas and making authoritative architectural decisions
Nice to Have
-
Experience with MLOps — GPU provisioning, model registry integration, inference pipeline deployment
-
Familiarity with Matrix protocol infrastructure (Synapse/Dendrite federation and worker architectures)
-
Experience with MongoDB, PostgreSQL, Neo4j, and Redis in cloud environments
-
Frontend deployment experience (CDN, static hosting, SSR)
