The AI Platform
The AI platform built for production.
TAIP unifies development environments, distributed training, model inference, and platform operations into a single stack — on-prem, in your cloud, or hybrid.
Builder products
Operator products
Cluster foundation
The suite
8 products
One control plane. One identity, quotas, policy, and billing across every tool your AI team touches.
GPU support
5 vendors
NVIDIA, AMD, Intel, Huawei Ascend, and Cambricon — auto-detected, no flags, no per-stack plumbing.
Deployment
Air-gap ready
Self-hosted identity, vendored charts, no required outbound dependencies. Same install online or disconnected.
Open standards · interoperates with the rest of your stack
Why TAIP
A platform — not another tool to integrate.
Most AI stacks are duct tape: one vendor for notebooks, another for training, a third for inference, a fourth for governance. TAIP brings them onto one plane, with one identity, one policy, and one bill.
One platform, the entire AI lifecycle
Notebooks, distributed training, model inference, and agents — on a single, consistent stack. Stop stitching together a dozen tools.
Built for production, on day one
Multi-tenant identity, quotas, policy, and audit are not an afterthought. TAIP is the platform you'd build if you had time to do it right.
Yours to run, anywhere
Install TAIP into your data center, your VPC, or air-gapped environments. Bring your hardware, your IdP, your data — TAIP fits in.
The TAIP suite
Focused products. One platform.
Each product is sharply focused. Together they form a cohesive plane for everything an AI team and the operators behind them need.
Everything researchers, engineers, and AI app teams need to ship — from a notebook to production inference.
ConsoleX
AvailableThe self-service Kubernetes workspace for every user
Each user gets an isolated namespace with quotas, storage, networking, and a web terminal — no kubectl, no tickets, no per-user RBAC.
Learn moreDevSpace
AvailableManaged AI development environments on Kubernetes
Single-click Jupyter, Marimo, Streamlit, Gradio, and VS Code environments — GPU-ready, isolated per user, idle-shutdown by default.
Learn moreTrainX
AvailableCurated, multi-tenant training on Kubernetes
Templates that describe themselves render directly into a UI form — admins control the script and defaults, users supply the parameters.
Learn moreInferX
AvailableSelf-hosted Model-as-a-Service for the agentic era
A unified gateway and admin plane for serving LLMs across providers and on-cluster runtimes — OpenAI- and Anthropic-compatible, OTEL-instrumented end to end.
Learn moreImageSphere
AvailableOCI registry with first-class identity, access, and ops
A curated fork of CNCF Sandbox Zot, extended with a real admin UI, runtime-mutable access control, OIDC SSO, namespace quotas, and a credible air-gap story.
Learn moreAgentX
Coming soonThe agent platform — built on TAIP
Build, run, and govern long-running AI agents on the same plane as your models.
Learn moreRun TAIP at any scale. Manage users, quotas, GPU pools, and policy from one console.
Stand up the underlying Kubernetes plane, identity, and storage that powers TAIP.
How it fits
Three layers. One stack.
TAIP is layered so each team owns the right thing. Builders use ConsoleX. Operators use TAIP Admin. IT owns TAIP Base. Everyone benefits from the same open APIs underneath.
-
1
Builder products
ConsoleX, DevSpace, TrainX, InferX, ImageSphere, AgentX — the surface AI teams actually use to ship.
-
2
Operator product
TAIP Admin — the control plane for the people running the platform: users, quotas, policy, billing.
-
3
Cluster foundation
TAIP Base — an opinionated, supported install of the Kubernetes, identity, storage, and networking plane underneath.
Builder products
Operator products
Cluster foundation
Solutions
Built for the teams that actually do the work.
TAIP isn't a generic platform — it's shaped around the day-to-day of research labs, AI product teams, and the platform engineers who keep them running.
AI research labs
From notebook to multi-node run, without leaving the platform
Give researchers GPU-ready dev environments, distributed training that scales, and a clean handoff from experiments to production — all under the same identity and quota model.
Read moreAI product teams
Ship LLM-powered products on infrastructure you control
Stand up OpenAI-compatible endpoints, host fine-tuned models, and run agents on dedicated capacity — with the cost, latency, and privacy your product needs.
Read morePlatform & infrastructure teams
Run a serious internal AI platform without building one from scratch
TAIP gives platform teams the multi-tenant control plane, policy, and observability that turn a GPU cluster into a real product for the rest of the company.
Read more