Guard Available
Review security evidence before use.
Scan public GitHub code or supplied source for risky permissions, behavior and other review signals.
Open Guard →Evidence, not a safety guarantee.
Model-agnostic AI runtime
AI that answers your customers, reads your documents and writes your content — on your data.
A model-agnostic AI runtime: Models, Memory, Skills, Agents and Workflows behind one secure API.GenCMS is the model-agnostic AI runtime direction behind customer knowledge assistants, document extraction, content engines and internal agents. Composer, Guard and tenant-scoped workspace records and APIs are available today; unified Skill execution, Memory retrieval and Agent execution remain on the roadmap. Try the available tools now, or scope one use case with iGears for a 30-day MVP.
Composer available now Guard scans available now Roadmap features clearly labelled
Build an AI Worker
DraftModel
AutoGPTClaudeGemini
Memory
Product KnowledgeBrand Guidelines
Skills
SEO AuditContent Writer
Output
AgentWorkflowApp
Usage estimate shown before each run
RunFour practical starting points
What people use GenCMS for
Start with one clear input, let GenCMS apply the right context and controls, and deliver a useful output.
Input
Approved knowledge
GenCMS
Retrieve with controls
Output
Grounded response
Input
Business documents
GenCMS
Read and validate
Output
Review-ready data
Input
Brief and brand context
GenCMS
Generate with guardrails
Output
Content for review
Input
Request and context
GenCMS
Plan within permissions
Output
Traceable next step
Each 30-day MVP is based on an agreed scope, data, integrations, acceptance checks and hand-over. Project services are fixed-price, with the fee and deliverables stated in the quotation.
Deployment options
Three ways to run GenCMS
Choose the model access, data location and level of infrastructure control that fit each workload.
01
iGears-routed managed API
Use one endpoint for supported leading global models, including selected models not officially open for direct signup in Hong Kong, without opening an overseas account.
02
Hong Kong data-centre hosting
Run a managed GenCMS environment in a Hong Kong data centre when local hosting and operational support fit the workload.
03
On-site
Run GenCMS on your own server. With local models and dependencies configured, the system can keep operating without an internet connection.
Deployment modes can be mixed by scenario.
Fit the tool to the job
Copilot and GenCMS
They solve different layers of work and can be used together.
Microsoft 365 Copilot
AI for each employee inside Outlook, Teams and Office.
GenCMS
AI for your own data, processes and channels, including websites, WhatsApp and ERP.
Use them together
Start with a free 30-day Copilot trial, then build the use case Copilot does not cover with GenCMS.
Frequently asked questions
What to know before you start.
Straight answers about GenCMS, deployment choices, data control and the first MVP.
- What is GenCMS?
- GenCMS is a model-agnostic AI workspace and runtime direction for Models, Memory, Skills, Agents and Workflows. Composer, Guard, tenant-scoped records, policy controls, activity metadata and scoped APIs are available today; Skill and Agent execution, Memory retrieval and the unified execution API remain on the roadmap.
- Who is GenCMS for?
- GenCMS is for developers, product teams and organizations that need custom AI, controlled access to business data or a private deployment. Customer-facing teams normally use an application such as AskCore, HumanLevelUP or PromoPilot, while GenCMS runs behind it.
- Does GenCMS lock us to one AI provider?
- The current tenant-facing contract is locked to AIHK in Hong Kong or AICA in Canada, selected by an administrator. Outside-provider switches such as OpenAI or DeepSeek, automatic routing and a unified multi-model execution API are not implemented yet.
- Can GenCMS run on private infrastructure?
- Yes. GenCMS can run through an iGears-routed managed API, in a Hong Kong data centre or on a customer's own server. These modes can be mixed by scenario, and an on-site setup can operate offline when it is configured with local models and local dependencies.
- How is GenCMS different from Microsoft 365 Copilot?
- Microsoft 365 Copilot helps each employee inside Outlook, Teams and Office. GenCMS is for AI built around your own data, processes and channels such as a website, WhatsApp or ERP; the two can be used together.
- How do we start?
- Developers can start with the existing Composer and Guard tools. Organizations can choose one bounded use case, data source and delivery channel for an agreed MVP that iGears can deliver within 30 days.
Available now
Start with Composer or Guard.
Use the tools that work today. Roadmap features are labelled separately.
Composer Available
Shape and test a model conversation.
Build messages and run chat through the tenant's region-locked AIHK or AICA gateway once its AI key is ready.
Open Composer →Workspace + roadmap
One place to assemble the AI behind your product.
Store tenant-scoped configuration today, then add execution and retrieval as the runtime roadmap ships.
Available today
Composer · Guard · tenant records · policies · activity metadata · scoped API keys
Roadmap
Skill and Agent execution · Memory retrieval · schedules · publishing · unified execution API
Build or deploy
Choose the next useful step.
Open the workspace, inspect the API boundary or choose where GenCMS should run.
01
Workspace
Configure one tenant and one bounded use case.
Open →02
API
See the live workspace surface and execution roadmap.
Open →03
Deployment
Compare managed, Hong Kong hosted and on-site modes.
Open →GenCMS is a product of IGEARS TECHNOLOGY LIMITED. Guard results describe observed signals, not guarantees.