InsightFusion
In-house telecom ops system: SSH/MML interrogation, anomaly detection, PCAP/Diameter comparison, and KPI reporting.
What it solves
Unifies multiple operations tools into one workflow: remote node interrogation, anomaly detection, config visualization, traffic validation, and reporting.
Focus
Fast root-cause: clear evidence, repeatable checks, and operator-friendly interfaces built for real outages and upgrades.
Capabilities
- Non-standard shell SSH login handling + MML command formation/execution + output parsing
- Meticulous anomaly detection with manageable sensitivity and categorization
- User interrogations across multiple network nodes (multi-hop operational workflows)
- PCAP comparison for Diameter AVPs (Gx/Gy), including unsorted message handling
- ECV/HCV vendor NE config visualization (Ericsson/Huawei) as interactive tree views
- Subnet calculator
- TMF priority trace log analyzer (Wireshark-like)
- Reporting: counters, KPIs, and operational summaries
Tech: Python, Django, JavaScript, D3.js, Prometheus, Grafana, structured logging.
Impact
- Reduced time-to-triage by automating SSH/MML interrogation and parsing outputs across nodes.
- Lowered alert fatigue using anomaly sensitivity tuning and categorization (actionable signal over noise).
- Saved tens of hours during investigations with PCAP/Diameter AVP comparison (Gx/Gy), including unsorted message handling.
- Accelerated config audits and change reviews with ECV/HCV tree visualizations for Ericsson/Huawei nodes.
- Improved day-2 operations with fast subnet calculations, trace-log analysis, and consistent KPI/counter reporting.
Why it matters
- Outages and regressions often hide in small deltas (AVPs, ordering, vendor config diffs). This system makes those deltas visible quickly.
- Non-standard access patterns (shell quirks, multi-step logins, MML workflows) are a reliability risk when done manually; automation makes results repeatable.
- Faster, evidence-driven investigations improve MTTR and make upgrades/migrations safer with clearer rollback decisions.
- Operator-grade reporting (KPIs/counters) supports capacity planning, SLA discussions, and proactive maintenance.
