The journey into the future of AI Network Experts begins here

Buckle up. You are about to unveil the secrets of modern AIOps Networks

SliceUp Ninja on fire, the ultimate Network AIops Weapon

SliceUp Dojo: Peak behind the curtains

Next-gen network operations are complex and require a data-centric & AI-first approach.

SliceUp Dojo is an AIOps platform that is the complex logic engine powering Sensei AI Network Expert. Its functionality goes beyond predictive analytics.

(1) Agentless data ingestion & Intelligent pre-processing

(2) Proactive Anomaly Detection

(3) Intelligent data routing

(4) Multi-dimensional event auto correlations

(5) Auto-remediation

(*) Digital Twin for “What If Analysis”

(1) Agentless data ingestion & Intelligent pre-processing

SliceUp automates data collection and onboarding, so you can enjoy its benefits right away.

  • Scalable Data Ingestion Through APIs
    supporting Syslog, SNMP, Synthetic Traffic, DPI, Kafka, and others

  • Real-time Log Parsing Automation
    dynamically identifies static and variable parts of logs. No more rigid regex/grok

  • L1/L2 Network Topology Auto-discovery
    through existing and/or proprietary SNMP collection systems

(2) Proactive Anomaly Detection

Once the data is onboarded, SliceUp analyzes it, looking for anomalies.

  • Over ten domain-specific ML models
    for Traditional and GPU/HPC Data Center as well as WAN use cases

  • The meta-model ranks the criticality of the incidents and prioritizes what matters most

  • "Human on the Loop" provides feedback to ML models the system can constantly learn and improve

(3) Intelligent data routing

SliceUp can automate data re-routing to different types of storage for cost optimization and remove human bias from the process.

  • >95% of telemetry data tells you: "Everything is fine". SliceUp forwards it to a cheaper storage option

  • <5% of data is abnormal. SliceUp detects it and forwards it for further correlations and analysis. 

  • No manual processes, no human bias. Let the SliceUp system decide which data is worth the hustle

(4) Multi-dimensional event auto correlations

Anomalous events are forwarded to SliceUp's auto-correlation engine, which understands contextual network relationships. It enables more accurate results for root cause analysis.

  • Event correlations across multiple dimensions like time, topology, sites, etc

  • Correlations across multiple telemetry domains and high fidelity data sources

  • "Human on the Loop" to provide learning feedback and adjust correlation strengths

Tab 1
Tab 2
Tab 3

(5) Auto-remediation

SliceUp distinguishes between anomalies and routine issues. Routine issues like duplex mismatch or cert expired  can be automatically remediated with Ansible scripts.

  • Streamlined remediation workflow
    with three levels of automation - silent, manual, and auto

  • Auto: Automatically opens a ticket and runs an attached script. Manual: Opens a new ticket and queues script for review and execution later

  • Auto-remediation workflow is triggered the next time the incident that matches the pattern is detected

Demo of end-to-end workflow. Example use case - an unusual spike in retransmission delay caused by duplex mismatch, fixed with automated remediation workflow.

Digital Twin for “What If Analysis”

Auto-discovered Network Topology. SliceUp integrates with existing SNMP data collection systems to auto-discover your network topology. By overlaying key metrics on an always current topology, network administrators can easily find where the future problems are.

Plan Ahead - Get Proactive: SliceUp allows you to withstand the next big product launch or a black swan event. Proactively mitigate the risk based on modeled scenarios and take the appropriate action - add additional links, increase bandwidth, or re-route traffic. 

Identify Resources Needed. SliceUp's "what if calculator" models different future scenarios to identify bottle necks and see which interfaces may become over utilized if traffic increases, adjusted for risk. Examples of metrics measured: Interface Utilization, TCAM, Memory.

Gif animation showing a SliceUp dashboard with "what if Calculator". AI for Network Performance “What If Analysis” to model future scenarios to identify bottle necks and see which interfaces may become over utilized if traffic increases, adjusted for risk. Examples of metrics measured: Interface Utilization, TCAM, Memory.

Ready for the next level?



How is it deployed? SliceUp is a Kubernetes-based, scalable deployment that can be installed in your Private Cloud or On-Prem. This way, your data never has to leave your environment. With our fully automated data preparation and parsing, it is designed to work well even in an air-gapped environment.

What are your existing out-of-the-box integrations? Some of the existing integrations include Syslog, Kafka, NetScout, Ansible, ServiceNow, SolarWinds, NNMI, and others. Contact us for a full list of integrations, or to discuss a custom integration to meet your needs.

How much historical data do I need for it to work? The system needs a minimum of a week's worth of historical data, but we recommend 30 days’ worth of data for it to establish the baseline for your environment accurately.