Private AI Workspace
Hong Kong Science Park Incubation Company · Founder-led Delivery · Hong Kong
Enterprise Private AI and Web3 Security Architecture Specialists
We build secure, controllable and practical next-generation technology infrastructure for companies that cannot afford data leakage or system failure.
One Ledger Limited helps enterprises deploy private large language models, build internal knowledge systems, and design reliable Web3 infrastructure. Projects are founder-led by Terry Ng, with 20+ years of enterprise IT, fintech, blockchain and technical delivery experience.
No hard selling, no vague concepts. The first call focuses on your needs, current environment, core risks and realistic next steps.
One Ledger Limited is a Hong Kong Science Park incubation programme company.
We focus on enterprise blockchain applications, private AI, blockchain infrastructure and digital trust technologies.
Security Review
Review architecture, APIs, wallets and contract interaction flows together.
Current Focus
AI Document Q&A
Operational Risk
New technology creates opportunity, but also new operational risk.
Companies want AI to improve document processing, knowledge search and internal efficiency, but contracts, client data, pricing, medical data and confidential business information should not be uploaded to public AI platforms.
Web3 teams need to ship quickly, but many lack senior technical governance across smart contracts, RPC nodes, wallets, APIs, permissions, private keys and infrastructure security.
Companies want AI, but data cannot leave the company.
Internal documents may contain contracts, pricing, client, finance, technical and operational data that should not be handed to public AI platforms.
Web3 teams need senior technical oversight.
Architecture, RPC, wallets, smart contracts, APIs, private keys and deployment workflows can create downtime, asset loss or reputational damage when handled poorly.
Core Services
Focused services for AI data control and Web3 architecture reliability.
Service 01
Private Enterprise AI Setup
Keep sensitive documents away from public AI platforms.
- Deploy Qwen, DeepSeek or suitable open-source models
- Run on company servers, private cloud, Mac Studio or PC
- Build RAG knowledge systems for PDF, Word, Excel and Markdown
- Show document citations and source references
- Configure user, department and document permissions
- Provide deployment documents, operating guides and training
Typical POCs can be completed in 7–14 business days, depending on scope.
Request Private AI AssessmentService 02
Fractional CTO & Web3 Security
Identify architecture, node, smart contract and operational security risks before launch, fundraising or enterprise partnership.
- Technical strategy, roadmap, architecture and vendor governance
- Solana, EVM, Hyperledger Fabric, RPC, indexer and wallet integration
- Wallet signing flow, API permission and private key review
- Smart contract interaction flow and code pre-check
- Monitoring, logging, single point of failure and vendor risk review
- Clearer technical credibility for investors and partners
Available as monthly advisory, one-off architecture review, security review or targeted technical advisory.
Request Web3 AssessmentService 03
ERP AI Application Development and Implementation
Practical AI workflows for ERP, CRM, WMS and internal operations.
- Review existing ERP / CRM / WMS data structures and workflows
- Build AI document search, quotation summaries, procurement support and service knowledge bases
- Design human approval workflows for high-risk actions
- Connect internal APIs, shared folders and reporting data where appropriate
- Set up permissions, logs, audit records and operating guides
- Support expansion from POC to production internal applications
For companies with existing ERP or internal systems that want AI to improve search, customer service, procurement, quotation and operations.
Discuss ERP AI ApplicationsPricing Indicators
Clear scope, explicit deliverables and pricing based on actual requirements.
Private AI POC
From HK$30,000
For companies that want to validate private models, enterprise document Q&A and real usage quality.
Apply for AI POC AssessmentProduction Private AI
Quoted after assessment
For production usage, multi-user access, permissions, integrations, backup, monitoring and ongoing support.
Discuss Production DeploymentFractional CTO Advisory
Monthly advisory plan
For teams that need senior technical leadership without hiring a full-time CTO.
Discuss Monthly AdvisoryWeb3 Architecture or Security Review
Quoted by review scope
For teams preparing launch, fundraising, enterprise partnership or major upgrades.
Request Technical ReviewProduct Portfolio
Owned products and demonstration applications
Beyond advisory and implementation services, One Ledger continues to build products that demonstrate physical asset digitization, NFC interaction and blockchain verification capabilities.
Infrastructure Product
Tappix
Tappix uses NFC and blockchain to build physical-to-digital infrastructure, helping brands create verifiable, interactive and traceable asset experiences.
- NFC tap-to-verify physical identity entry point
- Brand membership, missions, content unlock and re-engagement campaigns
- Applicable to RWA, retail, ESG, collectibles and product authenticity
Flagship Demo
Auraxtal
Auraxtal combines NFC, personal AI memory and blockchain NFTs in a smart crystal bracelet, showing how physical products can become evolving, verifiable and interactive digital assets.
- Tap once with NFC to activate the guardian experience
- Build personalized interaction through AI memory
- Preserve milestones and collectible stories with on-chain verification
Selected Case Studies
Concrete delivery examples
We translate AI and blockchain technology into deployable, testable and maintainable business systems.
Electronic Components Trading
Private AI document Q&A for an electronics trading company
Private LLM and RAG knowledge base deployed on company-owned infrastructure, supporting PDF, Word, Excel and technical specification documents with department-level permissions.
Trading, Wholesale and Customer Service
Private AI customer service assistant
An internal AI knowledge base that generates source-backed draft answers from approved company documents while keeping human approval in the workflow.
Enterprise Permissioned Blockchain
Three-node Hyperledger Fabric architecture
Three-node Hyperledger Fabric network with Raft orderer architecture, Certificate Authority, channel design, chaincode deployment flow, API and blockchain explorer.
Some case studies are anonymized due to confidentiality obligations and may omit client names, internal data and sensitive technical details.
Founder-led Technical Delivery
Led directly by a senior technical founder
Terry Ng has over 20 years of enterprise IT, fintech, blockchain and technical project management experience across banking digital platforms, university research, logistics systems, Hyperledger Fabric, RFID/NFC and Web3 security.
He has managed over HK$10 million in government-funded projects and participated in a USD 500K fundraising. Today, he focuses on private AI deployment, enterprise knowledge bases, and architecture and security governance for Web3 projects.
Enterprise Private AIRAG SystemsHyperledger FabricWeb3 SecurityRFID and NFCProject Governance
Why One Ledger
More than advice: we help make the solution deployable.
Hong Kong Science Park incubation programme company
One Ledger Limited is a Hong Kong Science Park incubation programme company focused on enterprise blockchain applications, private AI, blockchain infrastructure and digital trust solutions.
20+ years of enterprise IT experience
Experience across enterprise systems, banking, ERP, logistics, RFID/NFC, fintech, blockchain, AI applications and technical project management.
Business requirements to technical foundation
We consider permissions, workflows, infrastructure, integration, operating cost and real usage scenarios, not only models or smart contracts.
Founder-led engagement
Requirements analysis, architecture design and key technical reviews are handled directly by the founder.
Clear scope and delivery standards
Each project defines scope, milestones, deliverables, acceptance criteria and estimated timing.
Engagement Process
How we work
Free initial consultation
15 minutes to understand your business goals, systems, key problems and timeline.
Technical assessment
Review data sensitivity, hardware environment, architecture, key risks and feasible options.
Proposal
Define scope, technical route, deliverables, timeline, acceptance method and quotation.
Deployment and acceptance
Execute deployment, testing, documentation, training and delivery by milestone.
FAQ
Frequently asked questions
Can private AI run without internet access?
Yes. Depending on security requirements, the system can be designed to run fully offline, within an internal network only, or through controlled APIs to approved services.
Does every company need an expensive GPU server?
No. Small POCs may use existing PCs, Mac Studio machines or single-GPU workstations. Hardware depends on document volume, concurrent users, response speed and accuracy requirements.
Does RAG mean AI answers are always correct?
No. RAG improves relevance and traceability, but still requires proper document processing, permissions, prompts, model selection, testing and human review.
Will documents be sent to public AI platforms?
Private deployment can keep models, documents, vector databases and applications inside the enterprise-designated environment. Data flow and external connections are defined during project design.
Do you provide formal smart contract audits?
We provide architecture review, code pre-check, risk analysis and remediation advice. Formal audits with legal, insurance or institutional recognition can be coordinated with third-party audit firms as required.
Can we hire only Fractional CTO services?
Yes. We can focus on technical strategy, architecture governance, vendor management, code review, technical hiring and project supervision.
Can an AI POC always finish in 7–14 business days?
Typical POCs with clear scope, normal document formats and ready hardware can finish in 7–14 business days. Larger cleanup, complex permissions, integrations or hardware procurement require separate assessment.
Can you use our existing servers?
Yes. We assess CPU, memory, GPU, storage, network and operating system before recommending whether to use, upgrade or replace existing infrastructure.
Final CTA
Use 15 minutes to decide whether your project deserves further investment.
Whether you are exploring enterprise private AI or need senior technical governance for a Web3 project, we can start from the current problem, key risks and feasible next steps.