NetFlow, IPFIX, sFlow & traffic context

Know what is using the network—and where it matters.

When a link is busy, the next decision depends on the traffic behind the number. AmrudAi connects conversations, applications, sites, interfaces and QoS evidence to explain demand, investigate unusual behavior and plan capacity around the services that use it.

The application experience

Flow & traffic analytics

See the complete workspace. Enlarge any image to inspect its panels, navigation and evidence at actual size.

Flow & Traffic Analytics — illustrative product view with sample data. Illustrative product view · Sample data
Explain bandwidth use through exporters, applications, conversations and queue classes.

Find what changed on a busy link.

Follow the conversation behind the bandwidth graph.

AmrudAi connects NetFlow, IPFIX and sFlow records to exporters, interfaces, applications, sites and service dependencies. Top talkers and traffic trends show where to investigate; sampling and template quality explain how much confidence the evidence deserves.

The capability in detail

Turn traffic records into an operational explanation.

01

Flow collection across exporters

Ingest NetFlow v5/v9, IPFIX, sFlow and compatible vendor/J-Flow formats with configured exporter identity and observation context. Normalize source, destination, interface and direction so different devices contribute to a consistent traffic picture.

02

Top talkers and conversations

Explore the sources, destinations, protocol and port combinations that account for traffic in the selected period. Move from a fleet view to a particular conversation while retaining its site, interface and time context.

03

Application and service context

Use exporter-provided application metadata, service mappings and observed relationships to explain traffic in operational terms. Keep the origin and confidence of classifications visible so a protocol or domain label is not mistaken for deeper application inspection.

04

Site, subnet and ASN analysis

Compare site-to-site matrices, subnet and VRF traffic, cloud destinations and ASN relationships. Trace where demand crosses business or provider boundaries and identify the circuits carrying the affected services.

05

QoS and traffic quality

Place class-level offered traffic, queue pressure and drops beside utilization and the relevant conversations. Separate a growing bandwidth demand from a policy or congestion problem affecting a particular service class.

06

Historical identity and enrichment

Resolve addresses against the relevant time using inventory, DNS, DHCP and IPAM evidence. Preserve source, freshness and confidence for entity, cloud, ASN and authorized session enrichment so current ownership does not overwrite an incident's history.

07

Sampling and coverage you can assess

Expose sampling factors, template health, sequence gaps, exporter resets and collection loss. Where compatible, compare Flow totals with interface evidence to show the observed coverage and help engineers judge the strength of the analysis.

08

Traffic trends, anomalies and capacity

Compare current demand with prior periods and established baselines; inspect sustained utilization, percentiles and growth. Connect unusual patterns to topology, route changes and service impact before choosing a capacity or policy response.

09

Focused incident evidence

Attach Flow summaries, selected conversations and traffic snapshots to the investigation. When deeper verification is required, authorized packet evidence uses explicit filters and time or size limits, preserving a focused record for review.

From information to action

Explain a busy link before changing it.

  1. 01

    Choose the service and interval

    Open the affected circuit, interface or site and select the period of degradation.

  2. 02

    Identify the traffic change

    Compare conversations, application metadata, traffic classes and site matrices with the preceding or comparable period.

  3. 03

    Check coverage and consequence

    Review sampling and exporter quality, then correlate interface drops, path performance and affected services.

  4. 04

    Preserve the decision evidence

    Attach the relevant traffic snapshot to the incident or capacity review and validate the effect of any approved change.

Built around your operating environment

Fit the way your team operates.

Align exporter formats, sampling, observation points and collection capacity with the decisions your team needs to make.

  • NetFlow v5/v9, IPFIX, sFlow and compatible vendor formats.
  • Conversations, sites, subnets, VRFs, interfaces, ASN and service views.
  • Exporter template, sampling, sequence and collection-quality visibility.
  • Privacy-controlled enrichment and bounded packet evidence.

What can flow records reveal about traffic?

Does Flow identify every application and URL?

Application detail follows what the exporter and trusted enrichment sources provide. AmrudAi retains that provenance. Ports, domains or exported application names do not imply full URL visibility, TLS decryption or inspection of every application payload.

What if an exporter uses sampling?

Sampling factors and provenance remain part of the traffic record and its interpretation. Coverage checks, sequence gaps and template health help engineers understand when totals are estimates or when evidence is incomplete.

Can we investigate traffic from a device whose address changed?

Temporal identity mapping uses the relevant inventory, DHCP, DNS and IPAM evidence for the selected time. The investigation can therefore retain the historical association instead of substituting today's address owner.

Does an empty Flow view mean there was no traffic?

No. The view may reflect exporter configuration, missing templates, sampling or delivery loss. AmrudAi surfaces collection quality and compatible interface evidence so the absence of records is assessed in context.

See the workflow in your environment.

Explore the technology, service dependencies and operating requirements that matter to your team.