Netwell Enterprise · Private AI Infrastructure

Run your AI inside your office. Not on someone else's server.

We install private AI infrastructure on your premises, harden it to a published security standard, train your staff to use it safely, and maintain it month after month. So you can use AI without leaking your clients to a foreign cloud provider.

PDPA-aligned · Mapped to OWASP LLM Top 10 and NIST AI RMF · Senai, Johor
The problem

Why generic cloud AI is no longer a defensible answer

Your professional firm is running on AI whether you've planned for it or not. The question is whether the way you're running it can survive an audit, a breach, or a cross-border data transfer review.

Your client data is leaving Malaysia

Every prompt typed into ChatGPT, Claude, or Gemini is processed on foreign servers, subject to foreign law and foreign subpoena. If you handle confidential client information, that's an unaddressed PDPA exposure sitting in your day-to-day workflow.

§

PDPA 2024 raised the stakes

Mandatory data breach notification. Mandatory Data Protection Officer for qualifying organisations. Materially higher penalties. Tighter cross-border transfer rules. Quiet handling of an AI-related leak is no longer an option.

Your staff are already using it — without rules

Junior employees paste contracts, patient notes, and financial records into chatbots every day. There's no audit trail, no acceptable use policy, and no way to recall the data once it's left the building.

Our answer

What Netwell does about it

We install a private AI server in your office, configured to the Netwell Private AI Security Standard (NPAISS) v1.0, integrated with your document workflows, and operated under a service agreement that keeps it patched, monitored, reviewed, and accountable.

Your data stays on your premises unless you explicitly authorise it to leave. When it does leave — for example, when a query falls back to a cloud model — it's logged, classified, and screened for confidential content first. You see exactly what left and when.

See exactly what we install →
Process

How an engagement works

STEP 1

Assess

30-minute consultation, then a one-day site assessment. We document your data classes, workflows, and risk profile. You receive a written report whether or not you proceed.

STEP 2

Design

Hardware sized to your team, models selected for your use cases (including Bahasa Malaysia where relevant), integrations with your existing systems. You sign off the design before we order anything.

STEP 3

Install & harden

On-site installation. Network segmentation. Encrypted storage. Identity, guardrails, audit logging, egress controls. You receive the signed NPAISS attestation at handover.

STEP 4

Train, support, review

Staff and admin training. AUP signed by every user. Monthly health checks. Quarterly reviews. Annual incident-response tabletop. We're available when you need us.

What you receive

What's in the box

Every Netwell engagement, regardless of tier, produces a working AI deployment plus a documentation pack you can present to a regulator, an auditor, your insurance underwriter, or a tribunal.

The system

  • Private AI server installed and integrated
  • Local language models for your use cases
  • Retrieval over your documents (RAG)
  • Optional cloud fallback with DLP screening
  • Web chat, Microsoft Teams or Slack integration
  • Access controls, audit logs, alerting

The documentation

  • NPAISS v1.0 Attestation (signed)
  • Privacy Notice covering AI processing
  • Data Protection Impact Assessment (DPIA)
  • Records of Processing Activities (RoPA)
  • Information Security Policy — AI addendum
  • Incident Response Plan — AI-specific
  • Acceptable Use Policy (signed by every user)
  • Access Control Matrix
  • Backup & Recovery Procedure
  • Vendor & Dependency Inventory
  • Training Records & Attendance Log

Read our security standard before you talk to us

NPAISS v1.0 is the published baseline every Netwell deployment is built and attested against. 73 numbered controls across ten families, cross-mapped to the OWASP LLM Top 10, the NIST AI Risk Management Framework, and the seven principles of the Personal Data Protection Act.

It's public. Read it, share it with your lawyer, hand it to your DPO. If you're going to trust someone with your AI infrastructure, you should know exactly what they're signing up to deliver.

Built for

Who we work with

Netwell is built for Malaysian SMEs in regulated and confidentiality-sensitive sectors. If you recognise yourself below, we're talking to you.

Law firms

Privileged communications, draft contracts, due-diligence packs, court documents. Everything in your matter management system is exactly the data you cannot put into a foreign cloud LLM.

Medical clinics & specialist practices

Patient records, treatment notes, diagnostic correspondence. PDPA, the Medical Act, and professional ethics all converge on the same conclusion: don't upload it to a chatbot.

Accounting & tax firms

Client financial data, restructuring plans, tax positions, audit working papers. MIA professional standards expect you to know where this data lives.

Manufacturers with proprietary IP

Bills of materials, formulations, process recipes, supplier contracts. Your competitive position is in those documents. Don't train someone else's model on them.

Pricing

Three ways to engage

Indicative pricing — final scope depends on your environment. All tiers include the NPAISS v1.0 attestation pack.

Bronze
Foundation
Small offices, 5–15 users, low-to-moderate sensitivity workloads.
  • NPAISS v1.0 baseline build
  • ½-day staff awareness training
  • Standard documentation pack
  • Quarterly remote check-in
  • 30-day post-install support
From RM 25,000 setup From RM 800 / month
Talk to us about Bronze
Gold
Managed
Larger firms or high-sensitivity environments needing assured response.
  • All Silver inclusions
  • 24/7 monitoring
  • SLA-backed response
  • Annual full security review
  • On-call DPO advisory
  • Customised model fine-tuning
  • Dedicated account engineer
From RM 120,000 setup From RM 4,000 / month
Talk to us about Gold

Multi-year and pre-paid annual discounts available. Group arrangements through professional bodies (Bar Council, MIA, etc.) on request.

Trust

Aligned to the standards your lawyer and DPO already use

NPAISS v1.0 is cross-mapped to the major frameworks that govern AI security and Malaysian data protection — so the work we do can be evaluated against the standards your professional advisers already understand.

OWASP LLM Top 10
NIST AI RMF 1.0
ISO/IEC 23894
ISO/IEC 42001
PDPA 2010 (as amended 2024)
Cyber Security Act 2024
MITRE ATLAS
FAQ

Questions we hear often

Is local AI as capable as cloud AI like ChatGPT or Claude?
For the workflows most professional firms care about — document summarisation, contract review, research, drafting, internal Q&A over your own documents — yes. Modern open-weight models running on properly-sized hardware are very capable. For frontier-edge tasks, we configure an optional, audited fallback to a cloud model that you control.
How long does deployment take?
Bronze: typically 2–3 weeks from signed contract to handover. Silver: 4–6 weeks. Gold: 6–10 weeks. Most of the calendar time is hardware lead time and integration with your existing systems.
What happens if the model gets something wrong?
The model is a tool, not an oracle. Your staff are trained to verify before acting on AI output, especially for advice that goes to clients or regulators. The NPAISS standard requires this training and a signed Acceptable Use Policy before any user gets access.
Do we need our own IT staff to run this?
No. The Silver and Gold tiers include managed support that covers patching, monitoring, and incident response. You do need someone in-house designated as the operational owner — typically a DPO, an office manager, or your IT contact — but they don't need to be technical.
What if PDPA changes again?
Our standard is reviewed at least annually, more often if regulation changes materially. Existing customers receive an updated version of the standard and a re-attestation as part of their managed support, with no surprise fees.
What if our staff already uses cloud AI?
Many of our customers come to us in exactly that situation. Part of our staff awareness training addresses transitioning staff onto the private system, including practical guidance on what to never put into external chatbots. The Acceptable Use Policy formalises the new expectations.
Can you help us pass a customer / insurance / regulatory audit?
Yes. The documentation pack we deliver — including the signed NPAISS attestation — is designed to be presentable to exactly those audiences. We're available to attend audit sessions as a technical consultant where the engagement requires it.

Talk to us

A 30-minute call. We listen to your situation, answer your questions, and tell you honestly whether private AI infrastructure is the right move for you. No pitch deck. No commitment.

Email us →
Email: [email protected] YouTube: @netwellai Senai, Johor