Solution BriefFor IT Leaders · Data Teams · AI Engineers

From hardware to business answers in 10 minutes.

An AI server that answers your business questions in plain language. Wren AI's Agentic GenBI Layer, running on the on-prem Phison AI Data Platform.

Book a meeting →Try open source ↗

Wren AI · Agentic GenBI Layer

Answer business questions in natural language

Agentic modeSkillsMemoryGenBI app

Local models

GemmaQwenGPT-OSSReach out to our sales team to learn more about the models Wren AI supports.

30+Enterprise AI deployments

2× ↑ UpConcurrent users

5× FasterTTFT vs. recompute

→How it fits together

From your hardware to business answers.

People and AI agents

Ask business questions in plain language.

Agentic GenBI Layer

GenBI agent and governed context. Turns questions into trusted answers.

Wren AI

Phison AI Data Platform

Runs large models on smaller, on-prem GPUs by extending memory into NAND.

Phison

Your databases

Stay inside your network. No migration, no copies.

Wren AI is the Agentic GenBI Layer, running on the on-prem Phison AI Data Platform. One stack, from your data to a trusted answer.

02About Phison Electronics

Phison Electronics (8299.TWO) is a public company and a global leader in NAND flash controller and storage solutions. Phison's market cap exceeded USD $18 billion as of May 31, 2026.

With 300+ enterprise AI deployments via its aiDAPTIV+ platform, Phison brings proven infrastructure reach and enterprise readiness to the joint solution.

Phison's aiDAPTIV+ middleware and aiDAPTIVCache SSD extend GPU memory into SLC NAND, letting smaller GPU systems run models that would otherwise require a GPU cluster.

03About Wren AI

Wren AI is the open-source GenBI agent and agent-agnostic context layer for enterprise data, with 15k+ GitHub stars and recognition as a leading GenBI solution.

With deployments across healthcare, financial services, and technology companies, Wren AI helps organizations query data in natural language with governance, air-gap compatibility, and open-source transparency.

From on-prem hardware to business answers, in one integrated path.

Product highlights

01 / one-stop

From hardware to business answers, in one solution.

Connect your existing database. In 10 minutes, your team asks questions in plain language. No code, no custom build.

02 / on-prem by default

Break the GPU memory ceiling.

aiDAPTIV+ extends GPU memory into NAND, so smaller GPUs run large models. On-prem and air-gap ready.

03 / ai employee

An AI employee, not just a chatbot.

With tools, skills, and memory, it finishes tasks end to end and learns from your team, fully under your governance. It works, it does not just chat.

→Where it ships value

Three use cases, one platform.

01 · Replace traditional BI

Plain-language analytics, not dashboards-on-request.

Move from SQL tickets to plain-English answers in minutes.

Retail · E-commerce · Martech

02 · On-prem deployment

Compliant, physically isolated AI analytics.

Air-gapped, role-based, and auditable. A practical path to GenBI.

Healthcare · Finance · Government · Manufacturing

03 · Embed in the workflow

Bring analytics into Slack, Teams, and MCP.

One interface for analysts and MCP agents. No duplicated logic.

Ops · RevOps · Frontline teams

Book a briefing →

→What powers every workflow

The agent layer behind every use case.

Tools, skills, and memory run on your hardware, inside your governance boundary.

01 / agentic mode

More than a chatbot.

Agents reason over many steps in a sandbox: query data, build charts and dashboards, read PDFs, and save skills.

02 / skills & memory

Reusable workflows. Compounding context.

Skills are your agent's standard procedures, so output stays consistent. Memory learns from past work and turns daily use into company knowledge.

03 / agent-friendly

Native Git and file system.

Skills, memory, and models are all files: versioned, branchable, and reviewable. Agents read and write them directly.

04 / genbi app

Custom dashboards from a single prompt.

Describe what you need and get a polished, governed dashboard in one click.

→Customer voices

What leaders say.

“With Wren AI on Phison aiDAPTIV+, enterprise AI deploys in days, not months. It supports 10+ databases with no migration and hits an 80% query accuracy rate.”

Wei, LinCTO, Phison Electronics (8299.TWO)

“Wren AI powers our SaaS product DemandSense, so customers ask in natural language and build charts and dashboards in real time. Putting analytics in non-technical hands is a game changer.”

Anna KhamitovaCTO, Impactable (US)

“Wren AI changed how we decide. Real-time, natural-language insight lifted our decision speed by over 50% and made data a company-wide asset.”

Shasta HoCEO, Nextlink (TPEX:6997)

Bring your database. Ask a question.

Get your first answer in under 10 minutes, on your own schema.

Phison × Wren AI: a one-stop on-prem GenBI solution

Book a meeting →Try open source ↗

contact@getwren.aigithub.com/Canner/WrenAIPrivate briefings available on request

Solution BriefFor IT Leaders · Data Teams · AI Engineers01 / 02

Phison × Wren AI

From hardware to business answers in 10 minutes.

An AI server that answers your business questions in plain language. Wren AI's Agentic GenBI Layer, running on the on-prem Phison AI Data Platform.

30+Enterprise AI deployments

2× *↑ Up*Concurrent users

5× *Faster*TTFT vs. recompute

01The problem

Every AI server ships with the same gap.

Hardware vendors solve high-throughput AI compute. They do not solve the question every stakeholder asks the day after the box lands: How do I actually ask questions of my own data?

Why this partnership matters

Phison brings proven enterprise AI infrastructure and deployment reach. Wren AI adds the governed context layer that turns that infrastructure into a practical business-answering system. Together, the partnership closes the gap between owning AI hardware and getting trustworthy answers from enterprise data.

In regulated industries, cloud analytics is often not an option because data cannot leave the building.

02The solutionProduct highlights

From on-prem hardware to business answers, in one integrated path.

01 / one-stop

From hardware to business answers, in one solution.

Connect your existing database. In 10 minutes, your team asks questions in plain language. No code, no custom build.

02 / on-prem by default

Break the GPU memory ceiling.

aiDAPTIV+ extends GPU memory into NAND, so smaller GPUs run large models. On-prem and air-gap ready.

03 / ai employee

An AI employee, not just a chatbot.

With tools, skills, and memory, it finishes tasks end to end and learns from your team, fully under your governance. It works, it does not just chat.

03The stack

Hardware layerPhison aiDAPTIV+

GPU memory that scales into your SSDs.

aiDAPTIV+ middleware extends GPU memory into an SLC NAND SSD cache, so smaller GPU systems run larger models.

Extends GPU memory by up to 8 TB with aiDAPTIVCache SSD.
Run larger models on smaller GPU footprints, no cluster required.
Fully on-premises deployment for air-gapped environments.
Practical economics for enterprise teams outside hyperscale budgets.

Agentic GenBI LayerWren AI · GenBI

From data to answers in under ten minutes.

Open-source GenBI agent and agent-agnostic context layer that closes the analytics gap.

Auto-MDL builds the semantic model from your schema.
Natural-language queries return SQL-backed, governed answers.
agent-agnostic: analysts and AI agents share one interface.
RBAC, audit trail, and air-gap support built in.

Solution BriefProof & rollout02 / 02

04Where it ships value

Three use cases, one platform.

0101 · Replace traditional BI

Plain-language analytics, not dashboards-on-request.

Move from SQL tickets to plain-English answers in minutes.

Retail · E-commerce · Martech

0202 · On-prem deployment

Compliant, physically isolated AI analytics.

Air-gapped, role-based, and auditable. A practical path to GenBI.

Healthcare · Finance · Government · Manufacturing

0303 · Embed in the workflow

Bring analytics into Slack, Teams, and MCP.

One interface for analysts and MCP agents. No duplicated logic.

Ops · RevOps · Frontline teams

05What powers every workflow

The agent layer behind every use case.

Tools, skills, and memory run on your hardware, inside your governance boundary.

01 / agentic mode

More than a chatbot.

Agents reason over many steps in a sandbox: query data, build charts and dashboards, read PDFs, and save skills.

02 / skills & memory

Reusable workflows. Compounding context.

Skills are your agent's standard procedures, so output stays consistent. Memory learns from past work and turns daily use into company knowledge.

03 / agent-friendly

Native Git and file system.

Skills, memory, and models are all files: versioned, branchable, and reviewable. Agents read and write them directly.

04 / genbi app

Custom dashboards from a single prompt.

Describe what you need and get a polished, governed dashboard in one click.

06Seven-day rolloutDeployment · not a demo script

01Day 1 · AMConnect your database

Postgres, MySQL, ClickHouse, Snowflake, Redshift, DuckDB, plus 10+ more.

02Day 1 · PMAuto-MDL builds your model

Wren reads your schema. Data teams review joins, metrics, and access rules.

03Day 2–3Seed real business questions

Capture the questions finance, ops, and leadership teams actually ask.

04Day 3–5Tune with the business

Refine the context layer against trusted answers until accuracy clears the bar.

05Day 5–6Pilot with users

Turn on governance and validate answers, dashboards, and workflows with a live group.

06Day 7Roll out and connect agents

Launch to production and expose the same governed interface to people and AI agents.

Security and compliance, out of the box

All data stays on-prem with zero external transmission.
RBAC is enforced at the context layer, not only the database.
Every query can be audited with user, timestamp, and SQL traceability.
Air-gap operation works without external network dependency.
Open-source transparency helps security and compliance reviews move faster.
Designed for HIPAA, GDPR, and enterprise governance requirements.

Who this is for

Mid-market enterprises with serious data complexity
Healthcare organizations under HIPAA constraints
Financial services teams with residency requirements
Government and defense organizations that require air-gap support
AI engineering teams building agentic workflows
Data teams drowning in ad-hoc SQL requests

Bring your database. Ask a question.

Get your first answer in under 10 minutes, on your own schema.

contact@getwren.aigithub.com/Canner/WrenAIPrivate briefings available on request