The AI cloud platform for |

Simplify. Deploy. Scale.

How it works

Kubernetes Reimagined for GenAI

We've supercharged Kubernetes for LLMs and GenAI apps. Our Oasis operator optimizes complex workloads and manages clusters across clouds, saving you money.

AI/ML Ecosystem

HuggingFace
Langchain
LlamaIndex
PyTorch
Custom Apps

Features

Metric Collection
Predictive Autoscaling
Resource Optimization
Custom Scaling
RBAC
Multi-Cluster
Custom Metrics
Scaling Logs

Oasis

Operator

Monitoring

Dashboard
Billing

Kubernetes

EKS
GKE
AKS
Any Cloud
On-Prem
Private Cloud

Amazing Developer Experience

Okik CLI (Ocli) accelerates deployment, reducing time to market for your LLMs and GenAI apps.

Our streamlined deployment process enables rapid prototyping and iteration. With Ocli, you can quickly test, refine, and launch your AI applications, giving you a competitive edge in the fast-paced world of AI development.

GitHub Repo

Local Development

Configure and develop services locally with ease, ensuring smooth transition from development to production environments.

@service(
replicas=1,
resources={"accelerator": {"type": "A40", "device": "cuda", "count": 1, "memory": 4}},
backend="okik"
)
class SmolLLM:

Rapid Iteration

Easily start and manage your Okik server. Create and monitor services to ensure development environment mirrors production.

> okik server
INFO Starting: Creating services...
replicas: 2 | resources: {accelerator: {type: 'A40', device: 'cuda', count: 1, memory: 4}} | backend: k8
INFO Creating service SmolLLM...
INFO Creating YAML for SmolLLM...
INFO Server listening to http://0.0.0.0:3000

Containerization

Quickly build and package applications into containers, ready for deployment across any environment with a single command.

> okik build
INFO Starting: Building application...
Entry Point: main.py
Verbose: False
Force Build: False
⠹ Building Docker image

Universal Deployment

Seamlessly deploy applications to any cloud provider or on-premise environment with our versatile deployment options.

> okik deploy
? Select a YAML file to deploy: smollm-config.yaml
YAML Configuration: smollm-config.yaml (Document 1)
? Do you want to continue with the deployment? (Y/n)
Deployment in progress...
INFO Deployment successful!