Skip to content

Introduction

ESC — OBSERVATION-STATE CONTROL FOR MODERN TELEMETRY

Built for the streaming telemetry era.
Optimized for gNMI. Domain-agnostic by design.

Modern telemetry moves signals efficiently.
ESC decides when those signals should become actionable proof.

  • gNMI-compatible
  • Domain-agnostic
  • Occurrence-based
  • Observer-time structured
  • Proof-driven activation
  • Reduce default telemetry expansion
  • Reduce delayed proof cost
  • Reduce incident ambiguity windows

ESC does not replace gNMI, logs, traces, metrics, packet capture, or vendor observability platforms. It adds an observation discipline above them: when to stay minimal, when to expand, and when to return to baseline.

"gNMI moves telemetry. ESC controls when telemetry becomes actionable proof."
ESC bounded occurrence emission graph
  • Baseline observation remains bounded while deeper evidence is collected only when structured deviation appears.
  • Emission follows bounded occurrence activity, not raw traffic volume or permanent deep capture.

ESC is not another telemetry protocol.
It is a domain-agnostic discipline for controlling observation state across telemetry sources.


In network environments, ESC naturally complements gNMI.
gNMI provides efficient model-driven telemetry. ESC decides when observation should remain minimal, when scoped gNMI subscriptions or other diagnostic mechanisms should activate, and when they should close again.

What about gNMI?

gNMI is one of the strongest foundations for modern network telemetry.

ESC does not replace it.

ESC adds a control discipline above it.

gNMI can expose modeled state, stream selected telemetry, and deliver updates efficiently. ESC addresses a different operational question:

When should observation stay minimal, and when is deeper evidence needed to reach proof?

In a gNMI environment, ESC can act as an observation-state controller:

  • keep baseline observation bounded;
  • detect weak deviation through structured occurrence behavior;
  • activate scoped gNMI subscriptions, capture, logs, traces, or vendor-specific diagnostics only when structured deviation is signaled;
  • return to a lower-cost observation state when deeper visibility is no longer needed.

In short: gNMI moves telemetry. ESC makes it proof-driven.


Beyond gNMI

ESC is not limited to networking.

The same discipline can operate above different telemetry and event sources:

  • network telemetry, including gNMI and model-driven telemetry;
  • dataplane occurrence sources such as EDT;
  • logs, metrics, traces, and OpenTelemetry events;
  • industrial signals, PLC/CAN/OPC UA/MQTT sources;
  • robotic, embedded, cyber-physical, or proprietary event systems.

The common pattern remains the same:

bounded occurrence evidence → temporal structuring → deviation identification → conditional evidence escalation

gNMI is a powerful foundation in networks.
ESC is the observation-control model across domains.


Why this exists

For more than twenty-five years in network operations, one pattern kept repeating: when a complex system degraded, the hardest part was rarely collecting more data. It was knowing which fragment of data deserved attention first.

Engineers would reconstruct behavior from scattered clues: counters, adjacencies, logs, packet captures, routing state, diagnostic outputs, and timing symptoms. The work was skilled, but slow. During that delay, bridges stayed open, teams expanded capture windows, storage grew, and uncertainty became operational cost.

ESC comes from that practical experience.

It starts from a simple observation:

Making systems more visible is not enough if the system still cannot decide when deeper evidence is worth collecting.

At scale, observability must not only collect signals.
It must control when signals become actionable proof.


Why ESC matters, in one minute

  • For gNMI operators: add proof-driven control above existing model-driven telemetry.
  • For Network/SRE operations: reduce uncontrolled subscription growth, broad capture, and delayed correlation.
  • For CTO/VP Engineering: reduce ambiguity windows during incidents.
  • For finance/strategy: govern CAPEX/OPEX exposure caused by telemetry inflation and delayed proof.

ESC does not replace your telemetry stack.
ESC defines a discipline for deciding when deeper evidence collection should begin.

For engineers, that means fewer default expansions of capture, retention, and ingestion.
For leadership, it turns time-to-trusted-proof into an explicit CAPEX/OPEX variable.


Strategic Market Signals

Modern telemetry has already moved toward streaming, subscription, and event-aware operation. ESC builds on that shift. gNMI is the clearest network example, but the underlying problem is broader: how to govern when deeper evidence should be produced, retained, or closed across high-consequence systems.

Market Velocity Observability Landscape
$28.5B+
Structural Pain Global 2000
$400B
ESC Efficiency Architectural Gain
100x – 200x

Decisive Observability for High-Consequence Infrastructure

Observability is no longer only about collecting more.

It is about controlling when attention and proof should be engaged.

ESC is a methodological framework for using structured occurrence behavior to guide explanatory observation before degradation becomes fully visible.

It reframes observability-system design as a problem of observation-state control, not continuous exhaustiveness.


Conceptual Structure

ESC introduces a decision layer upstream of existing observability instruments.
It does not replace telemetry systems.
It defines how systems decide when deeper evidence collection should activate from structured occurrences.

ESC
Observation-State Control
Structured occurrence behavior governs when observation remains minimal, expands into proof, or returns to baseline.
EDT
Occurrence Source
Dataplane activity becomes bounded occurrence evidence, including signals not naturally exposed as gNMI objects.
SOBT
Observer-Time Structuring
Occurrence movement becomes comparable in observer-time structure, including weak deviation and absence patterns.
COSAT
Selective Materialization
Finite observation resources are selectively materialized without turning baseline observation into permanent deep telemetry.

The principle is to preserve sensitivity through bounded occurrence behavior while activating deeper explanatory observation only when structured deviation makes it operationally necessary. In network environments, that deeper observation may include gNMI subscriptions, packet capture, logs, traces, vendor telemetry, or EDT-derived occurrence sources. Beyond networks, the same control discipline applies above each domain’s native event sources.


Start Here


Resources & Strategic Papers

Foundations

Economic Strategy


Executive Distribution (PDF)

Secure, non-instructional versions for offline review and board circulation.