Skip to content
Access
Request access

Platform as a Service

Your AI workspace, ready to use.

A private workspace with everything a data team needs to build and run AI: notebooks, pipelines, training and serving. We run the machines, the updates and the security underneath.

Request accessHow it works

Who it's for

Data and AI teams who want to build, not administer.

  • Data teams in banks

    Build risk and fraud models on customer data that never leaves Nepal.

  • Universities and labs

    One workspace for a department, with fair GPU access for everyone.

  • Growing companies

    Get a full AI toolset on day one, without hiring people to run servers.

What teams use it for

  • Fraud and credit-risk models, from notebook to production
  • A research group sharing GPUs fairly
  • Nightly retraining of a forecasting model

How it works

Three steps.

  1. 01

    Sign in

    Your organisation gets its own private workspace.

  2. 02

    Build

    Notebooks, pipelines and training, on whole GPUs or slices.

  3. 03

    Ship

    Put models behind endpoints and watch them in the same place.

What you get

Platform as a Service, in detail.

  • Everything in one place

    Notebooks, pipelines, training, serving and data tools, ready when you sign in.

  • Private to your organisation

    Your workspace is isolated from other customers. Only the people you invite can reach it.

  • We run the platform

    Upgrades, scheduling, monitoring and security are our job. Your team works on models.

  • See use and spend

    Watch GPU use, job health and cost as they happen, before the invoice arrives.

  • Grows with you

    Start on a slice of a GPU and move to whole or reserved GPUs without changing tools.

  • Help in Nepal time

    Engineers in Kathmandu who know the platform because they run it.

Under the hood

The key numbers.

GPU
NVIDIA H200 NVL
Location
Kathmandu Valley, Nepal
Billing
Per second
Full specifications

Billing

GPU time and storage used, in NPR

Commitment: none, or a reserved term. Rates are being finalised; every request gets a firm quote in rupees.

How billing works

Questions

About platform as a service.

Do we need Kubernetes skills?

No. We run the cluster underneath; your team works in notebooks, pipelines and endpoints.

Which tools are included?

JupyterLab, Kubeflow, Airflow, Ray, Spark, MLflow, KServe, NVIDIA NIM, Presto, Superset and TensorFlow, among others.

Can other customers see our work?

No. Each organisation has its own isolated workspace.

Start with platform as a service.

Tell us what you want to run. It takes about a minute.

Request access

Other ways to run