Vishnu Yash Pandey

NetSense AI

A network-operations prototype connecting telemetry anomalies to the evidence and incident context engineers need to investigate them.

6
Evaluation profiles
5/5
Resolvable-scenario agreement
30 min
Sampling interval
01 — The problem

Investigating a network anomaly means connecting pieces that live in different places.

Network-operations exposure during my Tata Teleservices internship helped shape this problem framing; NetSense itself is an independently built prototype using synthetic/demo telemetry, not Tata production data.

Before Flow: Disconnected Incident Discovery
01Monitor telemetry
02Notice abnormal metric
03Investigate affected link
04Compare metrics
05Assess signal
06Find/create incident
07Determine ownership
08Continue investigation
02 — The product hypothesis

The product hypothesis

The idea is that surfacing anomaly scoring directly beside link context could reduce manual dashboard correlation.

After Flow: Unified Evidence Pipeline
01Network overview
02Flagged corridor/link
03Evidence
04Prediction/context
05Linked incident
06Timeline/evidence
07Engineer/context
08Human decision

Note: This is a product hypothesis, not a measured productivity result.

03 — How detection works

How detection works

The current detection system is deterministic and rule-based — not a trained model. Telemetry (latency, packet loss, jitter, bandwidth) sampled every 30 minutes across 6 links, with a 48-hour trailing window and a 12-hour forecast/context horizon, on synthetic/demo data.

Telemetry MetricWeightSampling Window
Latency40%30m intervals
Packet Loss30%30m intervals
Jitter15%30m intervals
Bandwidth Utilization15%30m intervals

Weighted score vs. threshold 0.22 — below is healthy/no strong live signal, at/above is attention/anomalous. A deterministic cutoff, not a learned boundary.

04 — Computed vs. authored

Computed vs. authored

Computed from telemetry
  • Latency
  • Packet loss
  • Jitter
  • Bandwidth
  • Anomaly score
  • Health state
Authored demo context
  • Prediction label
  • Confidence
  • Fault label
  • Incident severity
  • Engineer assignment

No telemetry-to-fault classifier, no dynamic assignment algorithm, no autonomous incident creation or remediation today.

05 — The failure

The failure

“A healthy score sat next to an active high-severity incident.”

Incident INC-395, Bengaluru–Hyderabad Core corridor: live anomaly score ≈ 0.16 (healthy/no strong live signal) while the linked incident record showed active, investigating, high severity, fault label “BGP Peer Dampening.”

The score wasn’t simply wrong — 0.16 was genuinely below threshold at that instant. A single point-in-time score can look healthy while the corridor is oscillating or intermittently unstable over time — a temporal blind spot, not a bad threshold. Kept visible rather than tuned away.

06 — Evaluation

Evaluation

Six designed telemetry profiles — bifurcation, loss degradation, saturation, stable, recovering, oscillation — a designed evaluation set, not a production benchmark.

Deterministic Agreement5/5 agreement

For scenarios where the expected t=0 state was resolvable (of the six profiles tested). This is rule-based detection agreement, not ML accuracy.

The oscillation profile scored ≈ 0.16 and classified healthy despite intermittent instability — the same failure pattern as INC-395.

13records
10 Consistent 2 Flagged 1 N/A

Consistency audit: 10 consistent, 2 flagged, 1 N/A of 13 records (7 predictions, 6 incidents). Kept separate from detection-agreement.

07 — Future direction

Rolling-window, variance-aware temporal scoring

Rolling-window, variance-aware temporal scoring — weighing recent behavior, variance and oscillation rather than a single snapshot. Future work; not yet implemented.

08 — Real-user testing

Real-user testing

Real-world testing was conducted with the prototype in a confidential network-operations setting, using company data. The underlying data, organization-specific findings, and detailed user feedback cannot be disclosed under an NDA.

Confidential real-world validation

Real prototype use in a live network-operations context provided feedback that informed the product framing. Detailed evidence from that testing is not publicly disclosed.

Public evaluation

The public case study instead uses synthetic/demo telemetry, so the detection behavior, evaluation scenarios, and limitations here can be inspected and reproduced by anyone reading this page.

09 — Value & limitations

Value & limitations

Potential value — reduced investigation friction, less context switching, better decision context — is a hypothesis, not a measured MTTR or productivity result.

Known Architectural Limitations
01Point-in-time scoring can miss oscillating instability
02Synthetic/demo telemetry, not production data
03Narrow signal coverage
04Confidence not a validated model
05Authored demo context fields
06No public real-user validation