AI across the products
Answers and drafts from your own data
across GlowScale, Orbio and SnowTrack
Staff have numbers explained, content drafted, incidents summarized and routine steps automated in the product they already work in. InterCloud AI works inside GlowScale, Orbio and SnowTrack with the same data and permissions as the person asking. On the website, it answers your customers' questions in Thai and English.
Capabilities
Answers, early notice of changes and help with routine work in every product
Assistant in the Products
Users ask in plain Thai or English, for example why margin dropped in March, which pages lost traffic or what changed before an outage, and receive an answer drawn from live data with the query and sources shown. The assistant respects the user's role, so Viewers see only what Viewers are permitted to see.
Insights & Anomalies
Be told what has changed without searching for it. Scheduled summaries and notifications report items such as an unusual purchase order, a slow-moving stock line, a page whose conversion fell or a service consuming its error budget. Each insight links to the records behind it.
Automation
Hand repeated work to the assistant and keep the decision. Such work can be turned into assistant actions, such as drafting a supplier reminder, preparing the month-end checklist, translating and tagging a new article or opening an incident from a cluster of alerts. Every action is proposed first and confirmed by a person.
Website AI Chat
Give website visitors an answer at any hour. The chat on itcsuite.com answers questions about our products and services around the clock, keeps the context of the conversation and hands over to a human agent when the visitor wants to talk to sales. The same chat can be deployed on sites built with Orbio.
AI Avatar
Serve customers and visitors by voice in Thai and English. The AI Avatar is a life-like presenter that listens and speaks in both languages, built on real-time avatar streaming and neural voices. It can be used for product walkthroughs, front-desk kiosks, onboarding and customer service, with content and tone defined by your organization.
Choice of Model Provider
Use only the models your organization has approved. Organizations connect the model providers they approve, hosted or self-hosted, and choose which assistants use which model. API keys are stored write-only on the server and are never shown again after entry.
Use Cases
What each department gains first
Finance
The assistant explains variances against budget, lists the invoices behind a rise in receivables and drafts the commentary for the monthly pack, working directly from GlowScale's ledger.
Sales & Operations
Sales teams receive early warning of stock-outs, a summary of a customer's history before a meeting and draft quotations prepared from price lists and previous orders.
Marketing & Content
Articles and product descriptions are drafted in Orbio, translated into every site language and supplied with metadata and alt text, and a weekly note reports which pages perform best.
IT Operations
Operations teams ask SnowTrack what changed before an incident, receive a plain-language summary for management while the team is still working and have the assistant draft the post-incident report.
Customer Service
Website chat and the AI Avatar answer common questions in Thai and English, collect the details a human agent needs and hand the conversation over with full context.
Management
Management receives a morning briefing across finance, sales, digital channels and systems health, with one place to ask follow-up questions without waiting for separate teams to report.
How It Works
Answers from your data within your permissions
The assistant does not keep a separate copy of your business data. Each question is answered by querying the product the user is in, through the same APIs and permission checks the user interface uses, and the result is shown with its sources.
Your data stays in your workspace. Prompts and answers are processed to serve the request and are not used to train public models; where a third-party model provider is used, requests are sent through our servers under contracts that prohibit training on your content. This is how InterCloud AI is designed and configured by default, and current details for a specific deployment are available on request.
Workspace Admins decide which providers and models are available, which assistants exist and who can use them. Self-hosted and on-premise model options are available for organizations whose data cannot leave their network.
- Answers come from live product data through permission-checked APIs
- Sources and queries are shown with every answer
- Customer data stays in the customer's workspace
- Prompts are not used to train public models
- Admins control providers, models, assistants and access
Questions
Frequently asked questions
Which languages does InterCloud AI work in?
The assistant, the website chat and the AI Avatar understand and answer in Thai and English, and the Avatar speaks both with neural voices. Language is detected per message, so conversations that mix Thai and English work naturally.
Can it see data a user is not allowed to see?
No, every question is answered through the same permission checks as the product's own screens. A Viewer in GlowScale gets Viewer-level answers, and a regional editor in Orbio gets answers about their sites only.
Which AI models does it use?
Workspace Admins connect the model providers their organization approves, hosted or self-hosted, and choose which assistants use which model. InterCloud can advise on the choice during implementation, and API keys are stored write-only on our servers.
Is our data used to train models?
InterCloud does not use your prompts, documents or business data to train public models, and third-party providers are used only under terms that prohibit training on your content. Customer data stays in the customer's workspace. For deployments with stricter requirements, self-hosted and on-premise models are available.
Can the AI take actions on its own?
Assistant actions, such as sending a reminder, publishing a translation or opening an incident, are proposed with a preview and carried out only when a person confirms. Automations you choose to run unattended are logged like any other user action.
See your own questions answered by InterCloud AI
Bring real questions from finance, marketing or IT and we will show how the assistant handles them inside the products, and how the Avatar handles them for your customers.