ARGIS-SOC Service
AI agents run security operations for mid-sized companies, 24/7.
A SaaS SOC that automates monitoring, analysis, notification, response and reporting.
- AWS Japan region
- Azure OpenAI
- Japanese-language support
- SOC in Tokyo, Singapore and Chengdu
In a nutshell
「An AI guard patrols 24/7 and acts and reports the moment something is wrong.」
Instead of shrinking an expensive staffed SOC, we rebuilt operations around AI. People step in only when a decision is needed.
Sound familiar?
Four common problems and how ARGIS solves them
Staff wear many hats and cannot watch all the time
ARGIS
AI agents watch logs and tickets 24/7 and keep first response going.
Too many alerts, no clear priority
ARGIS
Each event is scored and re-ranked by asset value and attack progress.
Separate tools, no overall picture
ARGIS
M365, EDR, firewall, VPN and cloud are correlated into a single incident.
A SOC seems too expensive
ARGIS
AI automates first response; expert support is bought only as needed.
Four strengths
Smart, fast, simple and affordable
Smart
AI agents on Azure OpenAI read across logs and explain reasons and next steps.
Fast
Monitoring through reporting runs in one automated flow.
Simple
Keep your M365, EDR and firewall. Connect the logs you need, step by step.
Affordable
A shared SaaS platform and AI automation keep fixed costs low.
Service overview
AI runs standard operations; people focus on decisions
Your environment
Log sources
- Microsoft 365
- EDR
- Firewall / VPN
- Cloud
- Servers
24/7
ARGIS SIEM (SaaS)
- Collect & normalize
- Correlate
- Create incidents
- Evidence & reports
24/7
AI agent team (first line)
- Monitor & score
- AI analysis
- Notify
- Runbooks
- Report
Expert second line (option)
- Weekdays 9–18 JST
- Deep analysis & tuning
You
Your contact
- Receive alerts
- Judge business impact
- Approve key actions
Built on our own ARGIS platform
The AI agent team, operation flow and safe automation
See how ARGIS — the AI SOC platform behind ARGIS-SOC — works, with real screens, on the product page.
AI threat hunting
Looking for signs that have not surfaced yet
Monitoring handles what was detected. Hunting looks for signs that have not become incidents yet.
- 01
Set the scope
Period, logs, assets, focus
- 02
Search all logs
Unusual auth, traffic, privilege changes
- 03
Test hypotheses
AI forms and checks attack hypotheses
- 04
Assess impact
Users, devices, time, data
- 05
Write the report
Findings, evidence, next steps
Plans
Three plans by size and log volume
Pricing is quoted for your environment. Expert second-line support (Human L2) is an option on every plan.
Starter
- Company size
- Up to 100 employees
- Daily log volume
- Up to 5 GB/day
- Standard log sources
- 3 types
- Standard use cases
- 10
- Reports
- Monthly
- Automated response
- Notifications & simulation
- AI threat hunting
- 2 / year
- Human L2
- Optional
Business
- Company size
- Up to 300 employees
- Daily log volume
- Up to 20 GB/day
- Standard log sources
- 6 types
- Standard use cases
- 30
- Reports
- Weekly & monthly
- Automated response
- Up to 5 runbooks
- AI threat hunting
- 4 / year
- Human L2
- Optional
Business Plus
- Company size
- Up to 500 employees
- Daily log volume
- Up to 50 GB/day
- Standard log sources
- 10 types
- Standard use cases
- 50
- Reports
- Daily, weekly & monthly
- Automated response
- Up to 10 runbooks
- AI threat hunting
- 6 / year
- Human L2
- Optional
Custom design (Enterprise) — let’s talk
Your environment is beyond the Business Plus limits. We will design a plan around your log volume, sites and current operations.
SLA
First-notification targets by priority
| Priority | Examples | AI first notification (24/7) |
|---|---|---|
| P1 Critical | Spreading ransomware, privilege takeover, suspected major data leak | Within 15 min |
| P2 High | Likely compromise, multiple assets, access to verify | Within 30 min |
| P3 Medium | Limited anomaly; impact not yet confirmed | Within 4 hours |
| P4 Low | Information, trends, improvement ideas | Daily or weekly report |
ARGIS SaaS monthly availability is 99.5% (excluding planned maintenance, force majeure and customer-side causes). Data is kept 90 days as standard (extendable to 365), and planned maintenance is announced 5 business days ahead. Human L2 is optional; without it, the time to a human response after an AI alert is not guaranteed.
Responsibilities
Clear lines between AI speed, human approval and your decisions
| Area | Flyingnets / ARGIS | You |
|---|---|---|
| Logs & platform | Receive, store, detect, ticket, AI operations (24/7) | Provide logs, access and current asset info |
| Notification | Automatic alerts, approval requests, records (24/7) | Name owners and contacts; receive and reply |
| Automated response | Run approved runbooks and record results (24/7) | Approve automation scope and key actions |
| Business decisions & recovery | Provide facts and recommendations (24/7) | Decide impact, shutdown, recovery, legal and disclosure |
| Human L2 | Expert work within purchased hours (weekdays) | Describe the request and provide information |
AI self-improvement follows approval rules. Changes are logged and key logic is reviewed and approved first. Major business or legal decisions, public disclosure and data deletion are never automated.
1-month free trial
Try it on your own environment for one month
Requirements: name an owner and a technical contact, connect logs within 5 business days and join a 30-minute final review.
Period
30 days from connection
Who
New mid-sized companies in Japan, once each
Logs
Up to 2 types, 3 GB/day total
Use cases
Up to 10 standard, email alerts
Expert support
3 hours (kickoff, mid, final review)
Automated response
Simulation and recommendations only
Reports
Weekly summary and final assessment
Cost
Free (commercial SLA does not apply)
Onboarding schedule
Live in about five weeks
- 1W0
Pre-check
Agree scope, log volume, contacts, goals
- 2W1
Design
Define connections, alerts, use cases, runbooks
- 3W1–2
Connect
Set up connectors and confirm intake
- 4W2–3
Tune
Tune noise, thresholds, asset value, templates
- 5W3–4
Acceptance
Test alerts, notices, reports, runbooks
- 6W5+
Steady state
Move to 24/7 AI operations with monthly improvement
Six reasons to choose us
Why companies choose ARGIS-SOC
AI that acts, not just watches
Runs the whole SOC workflow from monitoring to reporting.
Sized for mid-sized companies
Start from 5 GB/day; buy expert hours only as needed.
Keep your current tools
Connect priority logs first and keep your investment.
Explainable AI operations
Reasons, priority, alerts, approvals and results are recorded.
Japanese language, hours and hosting
Japanese alerts, Japanese experts, hosted on AWS in Japan.
Protects both IT and AI
Watch IT security and AI usage risk on one screen.
FAQ
Q.Do we need to connect every log from day one?
No. Start with key logs and add use cases and runbooks as you see results.
Q.What size of company is it for?
Designed for 100–500 employees and 5–50 GB of logs per day. Larger environments get a tailored Enterprise proposal.
Q.Do you cover nights and weekends?
AI agents monitor 24/7 and alert by priority. Expert support is a weekday option.
Q.Could the AI take major actions on its own?
Major actions run only after human approval (HITL). Only pre-approved low-risk actions run automatically.
Start with a free 30-minute call
Tell us your current tools, noisy alerts and the logs you want to connect, and we will propose a minimal setup.