Services
Our services fall into four lines of work. Each lists what you get and how it is priced; open one to see who it suits, who it does not, and common questions.
See all prices on one pageLINE bots and custom automation take repetitive support, admin and content work off your team, cutting cost and improving efficiency. We also build the internal systems, websites and online stores around them, and longer-term development can be outsourced by the week or month. Projects with a clear scope can be estimated in the calculator first.
Auto-replies, rich menus and broadcasts, connected to your orders and bookings
From rich menus and keyword replies on your LINE Official Account to backend-connected order lookups and bookings. Your customers in Taiwan are on LINE — your service should be too.
What you get
Turn LINE friends into customer records you can actually see
Tags, segments, interaction history, purchase history: build a CRM from LINE conversations and behaviour, then re-engage with segmented pushes and automated journeys.
What you get
LINE support only. For multiple channels, see Company chatbot
Focused on doing LINE support well: AI answers the common questions, recognises intent, collects order details, and hands over to a person when it cannot. If you also want a website chat widget and an internal knowledge base, that is the Company chatbot service.
What you get
Feed it what you know and it answers your customers and your staff
Not just LINE. A chat window on your website, your LINE official account and an internal knowledge base, all from one bot. It reads your existing site, PDFs, quotes and FAQ, answers what it can, hands over to a person when it cannot, and keeps a record so you learn what people actually ask.
What you get
Organize the workflow and standardize the process before adding AI automation
Automation depends on clear systems and standard processes. We first map the steps, data sources and responsibilities with you, then agree on input and output formats, decision rules, exception handling and human review points. Once that foundation is in place, we connect APIs or let AI operate Windows programs and websites without APIs, with multiple AI agents sharing data preparation, execution and checking. We make each step explicit and its results checkable before gradually expanding automation.
What you get
Articles, podcasts and short videos, produced automatically every day
Hand your content pipeline to AI: research, drafting and review, with a rewrite whenever a draft fails; once saved, it generates a cover image, podcast audio and a subtitled short video, and publishes them to your site, podcast feed and YouTube. We run this pipeline every day on our own blog and podcasts: from 1 January to 3 October 2026, about 210,000 page views and about 139,000 podcast downloads (Google Analytics).
What you get
Multi-agent: work that one agent cannot finish goes to a team of them
A complex task split across several agents: one gathers data, one decides, one actually writes to your systems, one reviews. Each agent only gets the permissions and tools it needs, inside an environment with spend limits and a full record of what happened.
What you get
Let AI do the work, but ask you before it acts
Once AI can edit records, submit orders or deploy code, the main risk is it acting on its own. We draw the lines: which actions need a human, what each agent is allowed to touch, how much it may spend in a day, and what it did. Our own console uses the same guardrails.
What you get
Admin panels, internal systems, and the plumbing between them
The internal system your staff open every day: orders, stock, scheduling, approvals, reports, reconciliation. We start from what you currently do in spreadsheets and on paper, and turn it into a system with accounts, permissions and an activity log.
What you get
Products, payments and shipping notifications, done in one go
Self-built or open-source-based shops: products and stock, Taiwanese payment gateways (ECPay, NewebPay), e-invoices, logistics, LINE notifications and a back office. We run an e-commerce business ourselves and know where it gets stuck.
What you get
Websites, tools and small systems with a scope you can describe
Marketing sites, campaign pages, sign-up forms, lookup pages, small internal tools. Tell us the pages and the features and we quote a fixed scope and schedule. For the bigger thing your staff open every day, see Business systems.
What you get
Mid- to long-term development capacity, billed weekly or monthly
Need a small team building continuously without hiring? We provide fixed capacity on a weekly retainer across frontend, backend, DevOps and AI, reporting in your project-management tools.
What you get
No need to build an engineering team first: an embedded consultant / FDE works with your team for fixed hours each week, and you can learn AI development one-to-one or work with us in other ways.
An embedded engineer, fixed hours every week, working alongside your team
Forward Deployed Engineer: a fixed number of hours each week inside your project, doing architecture, writing the critical code and helping your team adopt AI and automation. Suits startup projects whose specs are still moving.
What you get
One-to-one, or alongside your team: build something that ships, with AI
With AI coding tools like Claude Code and Cursor, we start from your own idea and work one-to-one until it ships; for teams adopting AI development, we can also consult and set up the spec, testing and deployment workflow with you. The focus is working with the AI: break the request down, write the spec first, let it change the code, understand what it changed, and trace problems when it goes wrong. The project you build is yours.
What you get
Sweat equity, joint proposals, training and talks
Anything not on the list above: co-developing products, technical advisory for equity, joint tenders, corporate training and workshops. Let's talk.
What you get
Mature open-source software saves licence fees, and we write the missing features as plugins or modules (for example Kong plugins in Lua). After rollout someone looks after it: version updates, security patches and code adjustments based on real use. Quoted on request.
Build your infrastructure on open source; we write the missing piece and look after it once it is live
API gateway (Kong, including plugins we write), file transfer (SFTP/SFTPGo), open-source remote desktop (RustDesk instead of TeamViewer), identity (Keycloak), content platforms (Strapi, n8n, WordPress), AI usage monitoring (Token Monitor, plus a web dashboard), uptime monitoring (Uptime Kuma, which can be turned into a multi-node high-availability cluster). We use these ourselves as well.
What we can set up or customise for you
See how much AI your team uses, and what it costs, on one screen
Your team uses Claude Code, Cursor, GitHub Copilot, Codex, Gemini and more, so usage is scattered across vendors and nobody sees a quota running out. We collect usage records from each computer with the open-source Token Monitor, set up a Hub that combines multiple machines, and build a web dashboard: today's and this month's usage and cost, a ranking by tool, cache hit rate, remaining quota and reset time per plan, and when each machine last reported.
What you get
Remote support through your own server, with no firewall changes on the computers you connect to
We set up an open-source RustDesk server on your cloud host. Helping colleagues remotely and reaching office or shop computers all go through that one server, like a private line of your own. The computers you connect to only need internet access: they connect out to the server, so there is no need for a fixed IP or for opening inbound firewall ports. When two machines can reach each other they connect directly; when they cannot, your server relays the session. Sessions are encrypted end to end, with no limits on seats or connection hours.
What you get
Open models for virtual humans, AI video, voice and music run on your own GPU server, so images, video and audio are generated on your machines with no per-use fees. Quoted on request.
Virtual humans, AI video, voice and music, generated on your own GPUs
We host open models on your GPU server: each virtual human gets a dedicated LoRA so the face stays the same across scenes and outfits, and shoots everyday clips or acts in short sketches lip-synced in its own voice. A topic, a news article or a slide deck becomes a narrated, subtitled video, and you can also train a custom voice, generate background music and transcribe recordings. Images, video and audio are generated on your machines, with no per-call billing.
What you get