Skip the AI dreams.
Start by solving your most pressing problem.
Too many customer inquiries, endless reports, or data scattered everywhere? We start by building an AI tool you can actually use in 4–8 weeks, then expand once you see results. You do not need a budget of over a million NT dollars.
AI projects we build
Each addresses a specific pain point, rather than offering a grand transformation plan.
AI customer service / Q&A chatbots
Let AI reply to messages on LINE, Slack, and your website. Repeated questions no longer need manual replies, and you are notified immediately when a customer shows interest.
Learn more →Enterprise knowledge base (RAG)
Organize information scattered across cloud drives, Excel, and documents into a knowledge base. AI cites the original documents in its answers. It can connect to customer service or HR / regulatory chatbots.
Learn more →Report / workflow automation
Automate order processing, reconciliation, and monthly reports. Let software handle repetitive manual compilation, so people only review results and exceptions.
Dashboards
Turn data from Excel, ERP, POS, and GA into dashboards business owners can understand. They update automatically, without waiting for someone to compile the data.
LINE group operations analytics
Monitor dozens of LINE work groups. AI classifies messages, assesses sentiment, and provides a summary every morning, notifying managers only when issues arise.
Learn more →Meeting minutes automation
Automatically turn recordings into transcripts with speaker labels and meeting minutes, then save them back to your company cloud drive.
Learn more →AI search visibility (GEO)
Check and improve how well AI search engines such as ChatGPT and Perplexity understand and cite your website.
Learn more →AI agents and internal tools
Delegate multi-step work to agents: look up information, organize it, and produce drafts for human review. We also build small custom internal tools.
Adoption consulting and training
Review departmental workflows and identify the most worthwhile task to tackle first. Teach teams to use AI safely and judge which data can and cannot be shared with it.
Types of systems we can build
These are technical areas we have implemented. Client names, scale, and results are not disclosed.
Data center infrastructure management (DCIM)
Deploy as a single executable, with no Windows Server or external database required. Read PDU, UPS, and switch data, with automatic alerts for anomalies.
Go / Vue 3 / SQLite / SNMP v2c·v3 / Redfish
Unified monitoring for multi-vendor equipment
Use the same probe to integrate SNMP, Redfish BMC, Modbus, and HTTP health checks. Monitor equipment from different vendors and receive alerts in one view.
Go / SNMP / Redfish / Modbus
Internal self-study platform
Courses and scenario-based quizzes, with AI-generated teaching images and Google sign-in. Suitable for turning experienced colleagues’ domain knowledge into training materials for new staff.
Next.js / React / SQLite / Google OAuth / Docker
AIDaily daily AI news
Automatically fetch AI news sources every morning at 7, summarize them in Traditional Chinese, categorize them, and generate archives and RSS. Runs unattended.
Python / Claude Haiku / Docker / cron
How we work together
Free AI check-up
Spend 30 minutes discussing your workflows and pain points to identify the most worthwhile task to solve with AI first.
Review and quote
Confirm scope, approach, timeline, and cost in 3–5 days. There is no obligation to proceed, and everything up to this point is free.
Delivery in 4–8 weeks
Depending on scope, we build a usable solution in 4–8 weeks, deploy it, teach you how to use it, and hand over the full source code.
Ongoing support
One month of free maintenance after delivery, followed by optional monthly advisory for ongoing improvements.
Ways to work with us and pricing
AI check-up
Free
A 30-minute conversation and specific suggestions for improvement. No sales pitch or obligation to proceed.
MVP implementation project
from NT$30,000–150,000
A usable AI tool or automation system delivered in 4–8 weeks. Scope and price agreed upfront, with source code handover and one month of free maintenance.
Monthly advisory
NT$18,000–38,000/month
Regular monthly meetings, with questions welcome anytime via LINE / Email. Ongoing technical support and feature iterations, like an external AI team. Flexible monthly terms; stop anytime.
NDA available. Invoices issued by ChibaKuma Studio Co., Ltd. (Taiwan Business ID (Unified Business No.) 61981389).
Frequently asked questions
Does AI adoption for a business have to cost over a million NT dollars?
No. An MVP implementation project for an AI customer service chatbot or report automation tool starts from NT$30,000–150,000, with delivery in 4–8 weeks. We focus on a small project first to validate the results. The check-up is completely free, with no obligation to proceed.
Can a company adopt AI without IT staff?
Yes. These are the companies we serve most often. Just explain the problem you face, and leave the technical work to us. We also design the interface so non-technical staff can use it.
Can we start with something small to try it out?
We strongly encourage it. Start with a PoC or small tool, see the results, and build trust before expanding. This carries the least risk for both sides.
What if we decide not to proceed after the free AI check-up?
That is completely fine. The check-up is free. We provide specific suggestions for you to assess, without repeatedly chasing you.
How do you protect data?
We can sign an NDA. Development environments have access restrictions, and sensitive data is encrypted. The system can also be deployed in your own company cloud account.
Who owns the code after delivery?
We hand over the full source code and operating instructions, with no contract lock-in. One month of free maintenance is included after delivery, followed by optional monthly advisory for ongoing improvements.
Related articles
Notes on problems and decisions from real projects, on our tech blog.
Find your first step in 30 minutes
“I want AI customer service.” “Can we automate reports?” “I am not sure what is possible.” Whatever stage you are at, you are welcome to talk with us.
Email us to book a check-up