A

Andromeda Tech Stack

Orchestration platform routing AI training and inference across distributed compute

Technology, Information and Internet San Francisco, California 11–50 employees Privately Held

Andromeda built an orchestration layer for AI infrastructure that matches training and inference workloads across global compute supply. The stack—Kubernetes, Terraform, PyTorch, JAX, TensorFlow, plus observability tools (Prometheus, Grafana, Loki, Datadog)—reflects a deeply operational platform focused on cluster provisioning and workload scheduling. Active projects cluster around supply-demand matching, supplier onboarding automation, and capacity planning, with clear pain points in scaling, utilization, and manual operational burden—priorities matching a company moving from single managed cluster to multi-supplier routing at scale.

Tech Stack 26 technologies

Core StackKubernetes Terraform Helm Ansible Python Go Prometheus Grafana Loki Datadog Linux C++ PyTorch TensorFlow Bash JAX NCCL eBPF

What Andromeda Is Building

◆Challenges

Scaling training and inference infrastructure
Supply-demand matching
Compute utilization
Supplier onboarding
Scaling rapidly
Complex infrastructure challenges
Manual operational work
Optimizing ai cluster performance
Improving scheduling efficiency
Reducing compute cost

▲Active Projects

Supply-demand matching platform
Compute supplier onboarding
Capacity planning processes
Kubernetes cluster provisioning for customers
Automation tooling for cluster deployments
Monitoring and observability implementation
Core platform component development
Automation and internal tooling creation
Reliability and observability enhancements
Profiling training workloads

Hiring Activity

Accelerating7 roles · 7 in 30d

Department

Engineering

5

Sales

1

Seniority

Senior

3

Mid

2

Manager

1

Company intelligence

Find more companies like Andromeda by tech stack, pain points and active projects

Get started free

About Andromeda

Andromeda provides orchestration infrastructure for AI training and inference workloads, routing compute jobs across a distributed network of providers and data centers. The platform abstracts supplier diversity and global capacity constraints, delivering compute when needed while optimizing for cost and utilization. The company operates in San Francisco with an 11–50 person team focused on engineering and infrastructure, currently in rapid hiring mode across senior and mid-level engineering roles.

HeadquartersSan Francisco, California

Company Size11–50 employees

Hiring MarketsUnited States

Frequently Asked Questions

What tech stack does Andromeda use?

Kubernetes, Terraform, Helm for infrastructure; Python, Go, C++ for application logic; PyTorch, JAX, TensorFlow for ML workload support; Prometheus, Grafana, Loki, Datadog for observability; eBPF and NCCL for low-level performance.

What is Andromeda working on?

Supply-demand matching, compute supplier onboarding, Kubernetes cluster provisioning for customers, capacity planning automation, workload profiling, and monitoring/observability to optimize AI cluster performance and reduce manual operational work.

Similar Companies in Technology, Information and Internet

Other companies in the same industry, closest in size

C

Candidata ai

Technology, Information and Internet

K

Kanz

Technology, Information and Internet

V

VIHU

Technology, Information and Internet