專人聯繫 免費試用
專人聯繫 免費試用
🏆 Among the first in Asia to run Agentic AI fully on-premises

Your AI base — your data never leaves, your AI stays yours.

The full SUPER 8 Studio (InsightArk + ORRA), deployed in your VPC or your own data center. Capabilities validated across 7B+ conversations, running under your environment's rules. The choice of finance, healthcare and government.

Book an architecture consultation → Enterprise AI deployment guide
7B+
conversations handled to date
1,000+
brands trust us
Taiwan's first
AWS AI Competency × Agentic AI dual designation

From a dedicated cloud to your own data center — you decide how deep the control goes

Conventional cloud AI asks you to move your customer conversations, member data and company knowledge onto the vendor's servers, on the vendor's terms. SUPER 8 Enterprise reverses that: the AI moves into your environment, and not a single line of data leaves. Three deployment modes, three depths of control — you always draw the line.

🔒 Dedicated VPC

Mid-to-large enterprises Brands with compliance requirements

An isolated environment of your own, fully separated from other customers. You specify where data resides, SUPER 8 helps operate it, and you keep full access control. The middle path between flexibility and control.

Book an architecture consultation Learn more
Architecture diagram of a dedicated VPC deployment

🏢 Fully private on-premises deployment (SUPER 8 MAX)

Finance Healthcare Government Organizations with the strictest data sovereignty requirements

The entire AI system installs in your own data center and can run fully offline. Models infer on your hardware, and conversation logs, knowledge bases and customer data live inside your perimeter from day one.

Book an architecture consultation Learn more
Architecture diagram of a fully private on-premises deployment

🔀 Hybrid architecture

Phased rollouts Multi-region operations

Sensitive data stays local, general workloads run in the cloud — you draw the line between what stays inside and what goes out, and you can move it whenever you need to.

Book an architecture consultation Learn more
Architecture diagram of a hybrid cloud and on-premises deployment

You pick the model. No vendor decides for you.

Which model powers your AI should be your decision, not a vendor's default. SUPER 8 Enterprise supports both commercial and open-source models: the model you run today can be swapped tomorrow — data portability, deployment autonomy, no lock-in.

Commercial models (cloud or VPC)

Anthropic, OpenAI, Gemini and other leading commercial models, configured freely by use case and cost.

Open-source models (can run entirely on your hardware)

GPT-OSS, DeepSeek, Gemma, NVIDIA Nemotron, Qwen — inference on your hardware, with not a single token leaving your data center.

One platform. Swap the model layer whenever you want.

Your AI strategy does not have to be a bet on any single model company.

Data sovereignty your auditors can actually see

For finance, healthcare and government, "your data is safe" cannot be a slogan — it has to be evidence. SUPER 8 Enterprise turns data sovereignty into something an auditor can read:

📍 Data residency

You specify where conversation logs, customer data and knowledge bases physically live — inside your VPC or your own data center. Where the data sits is stated in writing.

🔍 Access trail

Who accessed what, and when — logged in full and available for review. Internal governance and external audit read from the same record.

🧾 Model version tracking

Every response is traceable to the model and version that produced it. Model updates are recorded and reversible, so every AI decision has a paper trail.

📋 Compliance mapping

Mapping documents for ISO 27001, Taiwan's Financial Supervisory Commission (FSC) rules and the Personal Data Protection Act — Taiwan's GDPR-equivalent regime — so your compliance and security teams can complete internal review.

Three deployments, one table

One product, three depths of control. The difference is not how many features you get — it is who holds the data and sets the rules.

功能 / Feature
☁️ Cloud (SaaS)
Start free trial
🔒 VPC
Book an architecture consultation
🏢 On-Premises
Book an architecture consultation
Deployment location SUPER 8 cloud Your dedicated VPC Your own data center
Data residency Encrypted environment managed by SUPER 8 The cloud region you specify Entirely inside your perimeter, can run offline
Model choice Leading commercial models Commercial models + deployment in your region Commercial + open-source models running fully on your hardware
Operations Fully managed by SUPER 8 Operated with SUPER 8's help Led by your team, with updates and support from SUPER 8
Compliance and audit Standard security controls Access trail + proof of data residency Full audit capability: residency, access trail, model version tracking
Time to launch First AI employee live in as fast as 90 seconds Weeks (depending on environment) Scoped against your infrastructure; timeline provided at the architecture consultation
Best for SMBs, fast validation Mid-to-large enterprises, compliance-driven brands Finance, healthcare, government

Start in the cloud today, move to private deployment tomorrow — no vendor change, no starting over.

Also available through AWS Marketplace →

Already in use where data sovereignty matters most

🏦 Financial services

Customer financial data sits under the strictest regulation there is. On-premises deployment lets AI serve customers while conversations and personal data stay entirely inside the bank, with records ready for audit at any time.

🏥 Healthcare

Patient data leaves no room for compromise. AI handles inquiries and knowledge lookups inside the hospital environment, and medical data never leaves your system perimeter.

🏛 Government

Public-sector data that cannot go to public cloud does not mean AI is off the table. A fully offline private deployment brings Agentic AI inside closed networks.

🏭 Manufacturing

Process parameters, supply chain data and plant knowledge bases are core competitive assets. AI deploys on the plant floor, and not a single line of confidential data leaves.

⚖️ Legal

Case details, contracts and client data are highly sensitive. AI assists with retrieval, summarization and knowledge lookup inside a private environment, and privileged content never leaves the organization.

⚡ Energy

Equipment parameters, operational records and critical infrastructure data cannot leak. AI supports operations and decisions inside an isolated network segment, and core systems stay under your control.

Real results inside private environments

Control and results were never an either-or.

66%

of service volume handled by AI

AI takes the front line of customer conversations inside their own environment, handling 66% of service volume — the data stays in-house, and the coverage runs around the clock.

LUXCAT

5,500

physicians served with expert knowledge

An expert knowledge service for 5,500 physicians, running on a private deployment. Drug information queries are answered on the spot, and sensitive physician interaction records stay entirely within the company's perimeter.

A major pharmaceutical company

Already a SUPER 8 cloud customer? Upgrading does not mean starting over.

Every AI employee you have hired, every knowledge base you have built and every automation you run in the cloud moves into your private environment intact. Data portability is a product principle here, not an exception we make for you — SaaS to VPC, VPC to on-premises, one product, deeper control at each step, and everything you have accumulated comes with you.

Talk to an architect about your upgrade path → Compare deployment options

Frequently asked questions

Can't find your question here? Contact our live support or send us a message, and a specialist will get back to you shortly.

It depends on the deployment mode. VPC is typically measured in weeks. On-premises starts with an infrastructure assessment and a security review, and we give you a firm timeline at the architecture consultation. Worth noting: once the environment is ready, your first AI employee is generated in minutes, tuned and deployed within half an hour — what takes time is preparing the environment, not the AI.

It depends on the models you choose and your expected concurrency — running open-source models fully on your hardware requires GPU capacity, while the local side of a hybrid architecture is comparatively light. At the architecture consultation we provide a hardware specification recommendation for your use case, which you can assess against existing data center resources, or we can help you plan a procurement list.

No. You deploy the exact same platform that runs our cloud service — same product, same capabilities, the only difference being that it runs in your environment. That platform has been validated in production across 1,000+ brands and 7B+ conversations. From there it keeps improving inside your environment, on your knowledge base and your data: Build, Test, Deploy, Optimize, Govern — the whole lifecycle happens inside your perimeter, and the results only get closer to your business.

In VPC mode SUPER 8 helps operate the system, and update schedules are agreed with your IT team before they run. In on-premises mode your team leads operations and we ship update packages on a regular cadence — installable in an offline environment, with the model version and changes in each update fully logged. Whether to update, and when, is your call.

Yes, and it is the path many customers actually take. SaaS, VPC and on-premises are three depths of control on the same product — the AI employees, knowledge bases and automations you have accumulated move across intact, with no starting over. Data portability, deployment autonomy and no lock-in are design principles here: control deepens as your needs do, and what you have built stays yours.

Private deployment is quoted per engagement, based on deployment mode, selected modules and scale — there is no one-size-fits-all price list, because no two enterprises want the same depth of control. Book a 30-minute architecture consultation: we start with your environment and compliance requirements, draw the architecture with you on the spot, and then give you a transparent quote.

Your AI. Your Data. Your Rules.

A 30-minute architecture consultation, where we draw it out with you: how the full AI system runs inside your environment, and how not a single line of data leaves it.

Book a 30-minute architecture consultation Explore Enterprise AI