FLOWCUTTER for ISPs: Find the Cause Before Subscribers Call
Flow-level diagnostics, automatic AI-driven alert emails, and subscriber visibility help your team find the cause before customers call.
When a subscriber says the internet is slow, an uplink runs hot, or an edge router starts dropping traffic, the first graphs usually confirm that something is wrong. They rarely say why. Interface monitoring can show utilization; BGP can show the intended path. Neither normally identifies the customers, applications, source networks, targets, or flow pattern behind the event.
FLOWCUTTER fills that visibility gap. It is a passive network-diagnostics platform for WISPs, ISPs, peering environments, and data centers. It analyzes flow telemetry—NetFlow, IPFIX, or sFlow—alongside signals such as SNMP, BGP, DNS, and threat intelligence, so an operator can move from an alert to an evidence-based explanation.
Automatic AI-driven alert emails explain suspicious findings in plain language, identify the IPs involved, and recommend what to investigate or fix. Dedicated views for infected CPEs and peering traffic help the team identify affected subscribers and understand routing behavior before a problem becomes a wave of support calls.
What FLOWCUTTER does
Routers already see conversations moving across the network. Their flow exports describe metadata such as source and destination, protocol, ports, timing, and volume. FLOWCUTTER collects that data passively and makes it usable for both automatic detection and ad-hoc investigation. It focuses on flow metadata—who communicated, when, and how much—not the content of those communications.
- Investigates traffic at flow level. Drill into the endpoints, ports, protocols, applications, ASNs, and time window behind a spike or customer complaint.
- Detects known patterns automatically. Built-in templates cover volumetric DDoS patterns, DNS and NTP amplification, scans, brute-force activity, and unusual service-port behavior.
- Learns what is unusual for your network. Behavioral baselines build over approximately two weeks, helping teams distinguish a truly abnormal change from ordinary traffic variation.
- Surfaces exposure. Its vulnerability-scanning capability checks configured IP ranges for open ports, service versions, and known CVEs, with findings kept in an auditable overview.
- Keeps the team informed at the right pace. Urgent findings can trigger instant emergency emails; daily email updates bring recent activity to the operator’s attention, while periodic snapshots help track ongoing findings. The interface is Grafana-native, with dashboards, ad-hoc queries, and alert delivery through email, Slack, or webhooks.
Automatic AI-driven alert emails
FLOWCUTTER brings findings to your team automatically through AI-driven alert emails. When a flow or IP shows suspicious behavior, the email explains what was observed, why it deserves attention, and recommended steps to investigate or remediate it. Staff do not need to be cybersecurity specialists or search through dashboards to understand the initial finding.
For potentially infected customers, alerts identify the IP generating suspicious SSH activity or other known attacker patterns. This gives the operations team a specific source to investigate, with context and recommendations for addressing the activity before it causes broader problems.
FLOWCUTTER also watches your IPs for exposed services that can create security risks, including open DNS resolvers, publicly accessible SNMP, and SMTP services with unsafe configurations. The alert helps the team assess whether the exposure is intended and recommends corrective steps, such as restricting access, correcting configuration, or disabling an unnecessary service.
The CVE scanner checks configured servers and IP ranges for known vulnerabilities and informs your team when it identifies a potential CVE exposure. Findings give operators a starting point for validating the affected service and prioritizing patching or other remediation.
Why that matters to an ISP
For an ISP, the problem is seldom a shortage of monitoring. The real pressure is deciding what to do while subscribers are affected, the help desk is escalating tickets, and an engineer is trying to separate a local fault from a network-wide event. Flow diagnostics reduce the guesswork.
1. Shorter path from alarm to root cause
SNMP can tell you an uplink is saturated. FLOWCUTTER can help reveal whether the cause is a DDoS flood, a customer device with an exposed service, a sudden application shift, or traffic concentrated on a particular route. That context is what lets an operator choose a proportionate next step instead of making a broad, disruptive change.
2. Better DDoS decisions
A traffic surge alone is not enough to plan mitigation. An ISP needs to know the attack type, targets, dominant source networks, protocols, ports, and timeline. FLOWCUTTER is designed to expose that flow pattern so the team can investigate measures such as edge filtering, rate limiting, upstream coordination, or BGP-based mitigation with better evidence.
3. Fewer subscriber-impacting mysteries
Not every outage is a classic attack. A new open DNS resolver, an infected device, a routing change, a bandwidth-heavy application, or an emerging capacity issue can affect many customers before the issue is obvious in aggregate graphs. Flow history lets the team compare before and during an event, identify the affected segment, and preserve the evidence after the traffic returns to normal.
4. Useful security visibility without an inline device
FLOWCUTTER reads exports from existing flow-capable devices—or a passive SPAN/TAP copy through its probe. It does not sit inline, reroute traffic, or require agents on customer devices. That makes it a pragmatic addition to networks where forwarding stability and change control matter.
5. A clearer operational record
Historical flow data supports incident write-ups, customer conversations, capacity planning, and security review. FLOWCUTTER can also correlate flow behavior with SNMP and BGP context, helping teams explain both the impact and the path to a corrective action.
6. Automatic alerts with practical recommendations
AI-driven alert emails tell the team when a flow or IP appears suspicious, explain the finding in plain language, and recommend investigation or remediation steps. Operators receive useful context automatically, helping them move from detection to a targeted response sooner.
7. Earlier action on infected customers and exposed services
Alerts identify subscriber IPs generating suspicious SSH activity and other known attacker patterns, helping the ISP investigate potentially infected customers and protect its IP reputation. Monitoring for open DNS resolvers, exposed SNMP, and unsafe SMTP configurations also helps the team close service exposures that could be abused.
8. Visibility into known CVE exposures
The CVE scanner informs operators about potential known vulnerabilities on configured servers and IPs. This helps teams identify affected services, validate findings, and prioritize patching or access restrictions before those exposures become incidents.
Identify infected CPEs before subscribers call
A compromised customer router or another device at a subscriber’s premises can create problems beyond one connection. Abuse traffic can damage the ISP’s public IP reputation, and suspicious activity may indicate participation in a botnet.
FLOWCUTTER includes a preconfigured infected CPE dashboard that shows the specific subscribers associated with activity that is damaging IP reputation or may indicate botnet involvement. CPE means customer premises equipment. The dashboard gives the team a concrete subscriber to investigate instead of a general warning that something on the network looks wrong.
Operators can review the evidence, confirm the affected subscriber, and coordinate remediation before the problem spreads or customers start calling. Where addresses are shared through NAT or CGNAT, correlating the finding with subscriber records and NAT logs helps identify the right customer. Automatic AI-driven alert emails also identify IPs generating suspicious SSH activity or other known attacker patterns and provide recommendations to guide the investigation.
Peering Matrix and BGP visibility for routing decisions
For larger ISPs with peering and transit connections, knowing how much traffic crosses an interface is only part of the picture. Operators also need to see the actual path traffic takes through their peering partners and transit providers.
FLOWCUTTER’s Peering Matrix combines peering metrics with BGP data and displays the relationships in a flow chart. This brings routing context and observed traffic together, helping engineers understand how traffic is distributed across peers and transit providers and whether it follows the intended routing policy.
Operators can use that view to guide BGP routing optimization, investigate unexpected traffic paths, and assess the results of routing changes. They can also ask their connected AI about the peering metrics directly, using follow-up questions to explore the traffic relationships without manually searching through dashboards.
The value is in the follow-through, not just the alert
In practice, a useful network report should help an operator decide what to do next. FLOWCUTTER can send an instant emergency email when an urgent condition needs attention, then use daily updates to keep routine findings visible. The weekly security snapshots reviewed for this article add longer-running context. Each starts with a compact count of active, new, and resolved findings, then presents the findings in priority order.
- Context: whether the finding is new or ongoing, and the affected host or service.
- Evidence: a defined time window, estimated flow or connection volume, and the finding’s share of the relevant network activity.
- Interpretation: questions that help distinguish a legitimate scanner, monitoring system, NAT/CGNAT address, or expected service from a real problem.
- Action: concrete next steps such as examining destinations and ports, validating NAT logs, restricting public access with a VPN or ACL, rate-limiting, contacting a customer, or adding a known service to the network profile.
- Proof path: a direct Grafana link for drilling into the exact IP and time range behind the finding.
How it fits beside the tools you already use
FLOWCUTTER is not a replacement for every operational system. Its strength is providing network-wide behavioral context that those systems often lack.
| Tool or signal | What it answers | What flow analysis adds |
|---|---|---|
| SNMP | How much traffic crossed an interface? | Which endpoints, protocols, ports, and traffic sources created the load? |
| BGP | How should traffic be routed? | How is traffic actually moving, and what changed between peers or paths? |
| SIEM / IDS / IPS | What events or packet-level signatures were observed? | The network-wide scope, duration, affected systems, and communication pattern. |
Deployment and data control
FLOWCUTTER supports on-premise deployment for teams that want flow data to remain inside their infrastructure, as well as a hosted collector for a faster start. Existing routers simply export NetFlow, IPFIX, or sFlow to the collector; common supported vendors include MikroTik, Cisco, Juniper, Huawei, Arista, and Nokia.
FLOWCUTTER’s AI integration is read-only and works with the operator’s own approved AI tool. When connecting an external AI provider, use settings consistent with the ISP’s privacy and data-handling requirements.
A sensible way to evaluate it
Start with a real operational question: Which recurring problem costs your team the most time to explain? It may be noisy DDoS alerts, unexplained uplink saturation, customer complaints that cannot be reconstructed, exposed services, or uncertainty around peering changes. Export flow data from the relevant edge or aggregation devices, then validate whether FLOWCUTTER gives the team an answer that its current monitoring cannot.
FLOWCUTTER offers a 30-day trial. The platform provides visibility from the first day, while behavioral anomaly detection needs time to learn the normal pattern of the network. That makes the evaluation most useful when it includes both a current operational case and enough time to establish a baseline.
See what your monitoring does not show
If your team can see that a link is busy but not why, FLOWCUTTER can add the flow-level context needed for a faster, more confident response.
Talk with Link TechnologiesReferences
- FLOWCUTTER overview and ISP use cases
- FLOWCUTTER product capabilities
- FLOWCUTTER deployment options and sizing
- FLOWCUTTER ISP DDoS case study
About Link Technologies: Link Technologies helps ISPs and network operators design, operate, and secure resilient infrastructure.
IT Infrastructure • Wireless • Fiber • Hosting • Cloud Services
sales@linktechs.net • 314-735-0270 • shop.linktechs.net





Leave your comment