PHISON AI DATA PLATFORM

One integrated platform for more of your private AI.

The broader Phison AI Data Platform will bring applications, model operations, and infrastructure management together across servers. Run supported enterprise AI workloads on an integrated foundation, with hardware and software designed to work together.

GenBI is the first AIDP offering, available now on one server. The broader capabilities described here are coming soon.

One connected stack, top to bottom

The planned platform connects five layers, from hardware to enterprise agents. Each layer has a defined role. Applications use business data and model services; resource management coordinates the infrastructure beneath them.

Agent platform

Build, govern, and scale enterprise agents. 

Agentic hub

AI customer service

Document intelligence

Enterprise knowledge

Agent platform

Build, govern, and scale enterprise agents. 

Agentic hub

AI customer service

Document intelligence

Enterprise knowledge

AI data intelligence layer 

Powered by Wren AI. Connect business data and definitions 
to conversational analytics.

Data integration

Databases and data sources

Generative BI

AI data intelligence layer 

Powered by Wren AI. Connect business data and definitions 
to conversational analytics.

Data integration

Databases and data sources

Generative BI

Model and AIOps

Deploy, operate, and optimize supported AI models.

MLOps / LLMOps

Model management

Vision

Voice

Reasoning

Model and AIOps

Deploy, operate, and optimize supported AI models.

MLOps / LLMOps

Model management

Vision

Voice

Reasoning

Resource orchestration layer

Manage compute, memory, and storage across supported infrastructure.

Infrastructure management

Workspace and access

Operations and monitoring

Resource orchestration layer

Manage compute, memory, and storage across supported infrastructure.

Infrastructure management

Workspace and access

Operations and monitoring

Hardware layer

GPU compute, Pascari aiDAPTIV™ cache memory and middleware, and Pascari enterprise SSDs. 

GPU servers with aiDAPTIV

Cache servers aiDAPTIV SSD

Storage servers with Pascari SSD

Hardware layer

GPU compute, Pascari aiDAPTIV™ cache memory and middleware, and Pascari enterprise SSDs. 

GPU servers with aiDAPTIV

Cache servers aiDAPTIV SSD

Storage servers with Pascari SSD

What the broader platform will bring together

Coming soon. Deployment, infrastructure, models, memory, and monitoring in one platform. Less for your team to connect and manage.
Integrated deployment

Bring supported hardware, orchestration, model services, and applications together through repeatable deployment configurations. Your team still provides data access, business context, and validation.

Phison HCI will centrally manage compute, storage, and GPU resources across supported servers, with visibility into how capacity is used.

Deploy, monitor, and optimize supported models without assembling all of the model-management tools yourself.

Pascari aiDAPTIV™ adds a flash memory tier alongside GPU memory and system memory. Reusing eligible model context reduces repeated processing. Benefits depend on the model, workload, and configuration.

Track performance, resource utilization, and operating costs through a common management view, as supported by the release.

Choose supported hardware and software combinations for your applications. Configurations and validation results will be defined for each release.

PLANNED APPLICATIONS AND WORKLOADS

Built for more than one use case

GenBI is the first solution. More applications and model workflows are coming soon.

Generative BI 

Explore business data in natural language, powered by Wren AI. The broader platform will extend the infrastructure options around this first application.

Enterprise knowledge and RAG

Help teams find information in business documents and knowledge sources. Retrieval-augmented generation uses that information to provide context for model responses.

Agentic hub

Deploy and manage enterprise agents that retrieve information and carry out supported business workflows.

Document intelligence

Extract and interpret information from business documents.

Model inference and fine-tuning

Run supported models locally, with GPU resources and eligible cache reuse matched to the workload.

Custom enterprise AI 

Build and integrate applications for your business on supported platform services.

One platform, with more applications and workloads planned.

Plan how your
AI environment will grow

COMING SOON

The broader platform will add multi-server capabilities, additional applications, and expanded infrastructure management.
Discuss your expected workloads with Phison. Supported expansion paths, compatibility, and upgrade terms will be defined for each release.

Where does private AI fit in your roadmap?

Still have questions?

Talk to Phison about how the AI Data Platform can benefit your organization.

Talk to Phison about the broader platform coming soon, or start with the current GenBI offering on one server.