Digital economy operates on network, Prilink provides BI-agent, a business tool to fill the needs for intuitive network visualization for monitoring and verifiable network metadata for AI transformation. Engineered for simplicity, BI-agent is easy to use and starts quickly. It enables IT consultants and tech-savvy teams to easily:
- identify all network endpoints participating in business operations
- identify the IP, organization name and geo-location of malicious endpoints
- visualize endpoint activities through multi-dimensional network telemetries
- provide network metadata for AI transformation in telecom, transportation, contact center and more.
- streamline WFM tasks in sectors where metadata and employee workload are highly correlated.
Specifications
BI-AWS (BI-endpoint in AWS)
Network Traffic mirroring:
source: EC2 Linux kernel
destination: BI-AWS
bandwidth: 5 Gbps per BI-AWS, up to 16 EC2 per BI-AWS
EC2 setup: VXLAN script creates 2 tc queues in kernel. Regular traffic queue has higher priority. Mirrored traffic queue sends VXLAN packets to the IP address of BI-AWS in the AWS region chosen by users.
EC2 setup time: approx. 1 minute per EC2
BI-GC (BI-endpoint in Google Cloud)
Network Traffic mirroring:
source: VM Linux kernel
destination: BI-GC
bandwidth: 30 Gbps per BI-GC, up to 16 VM per BI-GC
VM Setup: VXLAN script creates 2 tc queues in kernel. Regular traffic queue has higher priority. Mirrored traffic queue sends VXLAN packets to the IP address of BI-GC in the GC region chosen by users.
VM Setup time: approx. 1 minute per VM.
BI-edge (BI-endpoint, on-prem hardware)
Network Traffic mirroring:
source: VXLAN, switch mirror port, network tap
destination: BI-edge
bandwidth: 4 x 2.5 Gbps or 2 x 10 Gbps per BI-edge, RJ45
BI-edge:
IP: dynamic or static IPv4
dimension: 170mm x 152mm x 53mm, weight: 1.57kg
power: max 45W
setup time: approx. 30 minutes
BI-win (BI-endpoint, on-prem, software)
Network Traffic mirroring:
source: VXLAN, switch mirror port
destination: BI-win
bandwidth: up to 2.5 Gpbs per BI-win
BI-win:
host: user’s Windows 11 pro PC or Windows server
host resources: 2 cpu core. 2G ram
BI-win OS: Debian 12 ISO running on Windows Hyper-V
BI-win interface: dedicated intel nic
BI-win IP: dynamic or static IPv4
setup time: approx. 30 minutes
YouTube https://youtu.be/NhABXx1juV0
BI-endpoint
- track up to 1 million IP sessions simultaneously
- generate metadata per session every 15-min
- multi-dimensional metadata include:
- IPv4, IPv6, tcp and udp ports
- bandwidth, speed, packet count, duration, hop/ttl
- tcp QoS: reTx, outOfSeq, sync failure, tcp flags, window size
- media QoS: rtp/srtp packet lost, jitter, latency, MOS
BI-agent
- analyze up to 1024 BI-endpoints in the Cloud and on-prem
- cache network metadata up to 31 days
- enable Web-UI for network traffic visualization and metadata download
Web-UI Network Traffic Visualization
- interactive 7-day multi-dimensional telemetries
- IP geo-locations and organization names
- most active endpoints in terms of BW, IP-session count, and time duration
- malicious IP identified by open source FireHOL
- traffic anomalies identified by 31-day metadata analysis
- multi-dimensional metadata record display
- capture network packet from BI-endpoint to Wireshark
Actionable insights using network metadata and AI
BI-agent provides 15-day network metadata in CSV or other formats for AI transformation. Screenshot at left shows how user interacts with Claude to uses different data visualization techniques to extract business insights.
Enhance WFM Forecasting and Scheduling
1. network metadata provide historical network traffic patterns to identify peak business hours and trends for WFM forecasting
2. in some sectors, network traffic is highly correlated to the employee activities, such as login, CRM update, customer services, task processing time and other workflows. Therefore network metadata are efficient to transform employee workload into unified analytics that streamline the task of WFM scheduling
Contact: sales@prilink.com 1-289-301-2330
