The team that built it stays with you through operations and improvement.
We do not stop at "build and done." We keep improving while your system runs. From monitoring to improvement development, the development team stays on to handle it.
Why the same team stays on through operations
Maintenance and operations means the team that built your system also handles everything from monitoring and incident response to ongoing improvement development.
No one understands your system better than the team that built it. Every handover loses context, slows response, and breeds knowledge silos. When the development team also handles operations, root-cause analysis and improvement both move faster.
To keep your system healthy over the long term, we catch anomalies early through monitoring and continuously improve based on operational data.
Four areas of support
We cover everything seamlessly, from monitoring to improvement development.
Monitoring
We monitor system uptime and performance to detect anomalies early.
Incident response
From first response and root-cause investigation through recovery. A dedicated BrSE (Bridge SE) is your point of contact for escalation.
Regular maintenance
We continue updates, security response, and regular reporting.
Continuous improvement
We propose improvements based on operational data and carry out additional development, including AI applications.
The mechanisms behind our quality (bug tracking, review, audit) are covered in Offshore Quality & Japanese Support →
Incident response levels
Standard response levels based on urgency. Specific figures are defined in individual contracts.
| Level 1: Blocker | Level 2: Critical | Level 3: Minor | |
|---|---|---|---|
| Status | System down | Major impact (system still usable) | Minor issue with no impact on core operations |
| Acceptance / first response | Within 30 minutes | Within 1 hour | Within 1 hour |
| Resolution target | 2 hours | 4 hours (determined through discussion) | Determined through discussion |
| Coverage hours | Every day | Mon-Sat | Mon-Fri |
Three support scopes
Monitoring only
We focus purely on monitoring uptime and anomalies, making system status visible first.
Standard maintenance
Beyond monitoring, this covers incident response, regular maintenance, and reporting, supporting stable operations.
Maintenance + continuous improvement
While operating your system, we also propose improvements and carry out additional development, continuing to grow your system.
For details on contract types and cost estimates, see Engagement models & team structure.
Frequently asked questions
How are service levels (response time, uptime) determined?
We define them as an individual SLA based on your system requirements. The standard is support during Japanese business hours; 24-hour on-call support is available as an optional contract.
How do you maintain the operational quality of predictive AI?
When predictive AI is deployed, we maintain accuracy through MLOps (model accuracy monitoring, retraining, and drift detection). This is used in areas such as manufacturing and demand forecasting.
Can we request maintenance only?
Yes. You can choose from monitoring only, standard maintenance, or maintenance plus continuous improvement, based on your requirements.
What is the general cost?
It is a fixed fee based on scope and the target system. Since it varies by project, we provide cost guidance individually (free quotes).
We are happy to discuss operations and maintenance, too.
Even if you are struggling with an existing system's operations, you can start by simply making its status visible. Consultations are free.
* The above are standard response levels (for reference). Specific acceptance times, response times, and uptime rates are defined as an individual SLA based on the target system and contract.
* Coverage hours and holiday support (such as Lunar New Year) vary by contract. 24-hour on-call support is available as an optional contract.
* MLOps is a framework for maintaining and improving machine learning models in production.
* The track records for EARTHBRAIN and Photoruction (approximately 6 people per month, approximately 300 person-months cumulative, 400,000+ projects, etc.) are based on information published by the clients and our own continuous support record.
* The contracting entity is the Japan-based corporation (Miichisoft Japan Co., Ltd.). A Japanese-language point of contact is available.