Stellar began in Hong Kong and has gradually brought its products and technology to a global audience. Today, we operate internationally while continuing to serve Hong Kong, exploring AI applications alongside local users, businesses and developers.
Hong Kong is where our story started. Its connections across languages, cultures and markets led us to ask, from the beginning: how can products developed here respond to needs around the world?
Our next chapter centres on AskAIs, alongside continued investment in our own M1 large language model, API services, and chip and hardware partnerships. We want to bring AI into more websites, software products and business workflows, helping teams across markets serve customers, build products and grow their businesses.
AskAIs: helping customers take the next step
Sometimes a customer leaves a product because a question goes unanswered: does this feature fit their needs, how do they get started, or who can help after a purchase? For businesses, cross-border merchants and independent developers, these questions are both service requests and opportunities to understand customers and build trust. AskAIs aims to put help where the need arises, so users can get support within a website or software product and continue exploring, using or purchasing it.
AskAIs combines AI customer support, live chat and a shared inbox. AI first handles questions using a business’s product information, instructions and knowledge base. When it cannot answer, or a request needs human judgment, the team can take over the existing conversation. Software users can seek help without leaving their current interface, while cross-border customers can communicate with the team in multiple languages. The team can also review the retrieved content and checks in AI trace to understand the basis of an answer and identify gaps in the knowledge provided.
The value of a human response should extend beyond one conversation. Teams can organise useful answers and add them to the knowledge base, giving future similar questions a better source of information. This gradually preserves experience that would otherwise sit with individual staff members. AI handles everyday questions, people resolve exceptions and complex requests, and the team updates the knowledge that supports future service.
Websites and apps can integrate the AskAIs chat interface, with Telegram integration also available. Mac, iPhone and iPad apps are listed on the App Store, the Android app is listed on Google Play, and Windows has a standalone download, helping teams follow up with customers across devices. Current features and downloads are available on the AskAIs website.
Next, we plan to improve knowledge-based answers, human handover and service across channels, and explore connections with orders, accounts and business systems. Helping resolve more practical requests, with the necessary authorisation and security validation, is a future research direction. Our longer-term goal is for every software product to have a service interface that understands the product and responds to users’ needs, helping businesses reduce customer loss and improve opportunities for conversion and revenue.
M1: bringing intelligence into more products, beyond a chat interface
AskAIs is a ready-to-use product. Stellar’s own M1 large language model and its open API serve developers and partners who want to build their own products. Many teams already have software, customers and established workflows; they need a model service they can integrate into them. Through M1, we aim to let these teams retain their product design and add AI where it can be useful.
For example, software could offer assistance around the work a user is doing, while business systems could answer questions and organise information using their own data. Such applications need model services, but also require developers to incorporate business information, access controls and result validation. The M1 API provides a way to access the model; specific capabilities and usage conditions are governed by the official API documentation. For our ongoing long-term partners, we will continue supplying model and foundational technology services under existing arrangements.
Future M1 research will address practical applications: understanding different languages and business contexts, exploring collaboration between models and tools, improving inference efficiency and service reliability, and evaluating response times and usage costs. These are research objectives that require progressive validation. We want more developers to turn ideas into usable, maintainable products, and will explore closer links between model research and product needs. Read more in our M1 and model API introduction.
Chips and hardware: bringing research into practical applications
AI needs computation, but it also needs continuous access to models and business data. How data moves between storage, memory and computing units affects real-world performance. We therefore plan to make AI data access and engineering integration important research areas in our chip and hardware collaborations, examining how foundational technology fits software and its use cases.
As a foundational technology researcher and supplier, Stellar will work with third-party partners to define requirements, develop technical designs, and carry out integration testing and application validation. Projects may involve technology supply, joint research or product integration. The objective is to solve specific engineering problems and gradually bring research into practical products.
Products and partnership updates will be announced as work progresses. Specifications, responsibilities, availability and access will be explained in each formal release. Visit our chips and hardware introduction to learn more about this direction.
Starting in Hong Kong, growing alongside different markets
Stellar’s direction is to make product, model and hardware research respond to real needs. We will continue serving Hong Kong while learning from international use: Hong Kong’s multilingual and cross-market needs can help improve our products, while feedback from global users and partners can also support businesses and developers in Hong Kong.
We will share the results of this work through concrete product updates, improvements to model services and progress on partnership projects. To discuss product use or technical collaboration, please contact Stellar.