Circuit Stitcher

Inspiration

The most dangerous moment in global network maintenance is the "Execution Gap." After a latency anomaly is diagnosed, a Network Operations Center (NOC) engineer must manually translate a technical instruction into a physical reroute. I witnessed this daily: engineers switching between a Jira Service Management ticket and a complex geospatial UI to redraw paths.

This manual context-switching is the primary cause of Shared Risk Link Groups (SRLGs)—where two "redundant" paths unintentionally share the same physical bridge or conduit.¹ In the 2026 inference-driven economy, a single manual error here creates a permanent "Recompute Tax" on the global GPU grid, wasting (\approx \$4,800/\text{mo}) per GPU due to latency-induced cache misses.²

I built Circuit Stitcher to close this gap, transforming a fragmented manual process into a policy-bounded, autonomous execution loop.

¹ AI for Network Resilience Assessment ² Beyond the Topology Illusion; Tensormesh (2025), Inference Economics Research

What it does

Circuit Stitcher is the "Execution Co-Pilot" of the Aegius Framework — an Autonomous UI Navigator that becomes the engineer's hands on the screen. It ingests the diagnostic output from Node 1 (Latency Lens) and the strategic context from Node 2 (Situation Intelligence Brief), then physically executes the required circuit reroute across production geospatial interfaces. Note In case of a path that does not follow architectural policy e.g. a path through improbable mountaineous terrain. It badges in and points it out and offers a potential correct path way.

Aegius is a multi-node AI agent architecture designed to close the communication gap between infrastructure telemetry and executive decision-making. Each node serves a distinct function—from raw signal analysis to strategic narration to physical execution—operating as a coordinated system. The nodes were developed and submitted independently across multiple hackathon categories:

Node	Role	Hackathon Category
Latency Lens	The Vision Auditor	Live Agents
Situation Intelligence Brief	The Strategic Translator	Creative Storyteller
Circuit Stitcher	The Execution Co-Pilot	UI Navigator (this submission)

It interacts with production interfaces the way a human engineer would—visually. Through policy-bounded autonomous execution, it:

Bridges the Context Gap: It visually reads the execution payload—Origin, RTT target, and recommended path—directly from an open ticketing system in Jira Service Management. No scraping. No integrations. It reads the screen the way a human engineer would.
Executes Physical Rerouting: It navigates to Google Earth and physically performs the mouse movements and polyline stitching required to redraw the circuit along the policy-compliant path.
Enforces the Absolute Override Protocol: It is not a blind macro script—it is an Interactive Co-Pilot that constantly monitors the engineer's manual UI inputs. If a human engineer attempts to manually draw a fiber route through prohibitive terrain—such as a mountain range that would degrade latency—the agent proactively barges in via voice, blocks the engineering violation, and takes back control of the UI to execute the correct path.

How I Built It

I built Circuit Stitcher as the tactical execution layer of the Aegius Control Plane:

Google Gemini 2.5 Flash: Powers the visual UI understanding, translating visual "intents" into executable action tags.
Aegius Control Plane: A secure FastAPI orchestration layer deployed on Google Cloud Run that manages the real-time command relay between the browser and the Gemini Live API.
Visual-to-Action Mapping: I utilized a custom local runner to translate Gemini's visual coordinates into physical OS-level mouse clicks and keyboard events, enabling true UI navigation.
Hierarchy of Truth: The agent treats the RTT telemetry and SRLG policies as absolute truth, overriding any manual input that contradicts the physics of the fiber route.

Challenges I ran into

The primary challenge was solving the NP-hard problem of hidden Shared Risk Link Groups. Traditional algorithms struggle to identify every physical intersection as the network scales. I had to pivot from static scripts to "Geospatial Reasoning," teaching Gemini to recognize topographical boundaries as physical constraints.

Technically, synchronizing high-resolution screen frames with the low-latency command stream through Cloud Run relay buffers required meticulous tuning to ensure the "Barge-in" felt instantaneous during a manual human error.

Accomplishments that I'm proud of

99% Audit Velocity: I collapsed a multi-week manual investigation and stitching loop into a 10-minute autonomous sensing cycle.
SRLG Mitigation: I successfully demonstrated a "Co-Pilot Override" that prevents the unintentional creation of shared physical risks, protecting against downtime costs that average $33,333 per minute.¹

- Vision over API: I proved that visual UI navigation is superior to API-based automation for legacy infrastructure tools that lack modern developer endpoints.

¹ Beyond the Topology Illusion; Pingdom Industry Downtime Benchmarks

What I Learned

I learned that in the 2026 token economy, Interaction Velocity is a business moat. Delays exceeding 500ms are documented to erode user engagement by 30%, making real-time diagnostics and execution a revenue protection strategy.¹

I also discovered that the "Barge-In" isn't just an interaction feature—it is a critical safety net for autonomous UI navigation, turning a reactive tool into a proactive collaborator.

¹ Sparkco (2025), Interaction Velocity Research

What's next for Circuit Stitcher

🔗 Aegius Framework

The Aegius Framework is a three-node intelligence system submitted across the three Gemini Live Agent Challenge categories:

Node Role Category

Latency Lens The Vision Auditor Live Agents

Situation Intelligence Brief The Strategic Translator Creative Storyteller

Circuit Stitcher The Execution Co-Pilot This submission

Each node is self-contained, but together they form the Aegius Planetary nervous system—a governance-first AI framework for planetary-scale network resilience.

Circuit Stitcher is the third and final node of this larger vision. Together with Latency Lens (The Vision Auditor) and Situation Intelligence Brief (The Strategic Translator), it forms the nervous system of distributed intelligence—three specialized nodes designed to sense, translate, and execute across planetary-scale network infrastructure.

These nodes are governed by the overarching Aegius Planetary policy firewall logic, ensuring every automated diagnostic, briefing, and execution step is grounded in hard operator metrics, enforcing AI Sovereignty and IP protection across the infrastructure grid.

Circuit Stitcher sits at the most consequential point in this chain. Where Latency Lens senses and SIB translates, Circuit Stitcher acts—making it the node where governance failures carry the highest cost. The roadmap is structured in four phases, ordered by complexity.

Phase 1: Agentic Execution

The immediate priority is to evolve Circuit Stitcher from a single-task co-pilot to a multi-agent execution swarm:

Zero-Click Pipeline Completion: Completing the automated handoff where Latency Lens diagnostic outputs flow through SIB's translation layer and arrive at Circuit Stitcher as structured execution instructions—closing the Sense → Translate → Stitch loop without manual intervention.
Agentic Swarm Orchestration: Scaling to a modular architecture where specialized agents handle parallel sub-tasks—topology validation, SRLG conflict detection, terrain analysis, and policy auditing—before any physical reroute is committed. Each sub-agent operates within bounded autonomy and reports to a central orchestrator.
Expanded Override Vocabulary: Extending the Absolute Override Protocol beyond terrain violations to detect and block a broader set of engineering errors—including SRLG conflicts, latency budget overruns, and regulatory boundary crossings—each triggering a context-specific voice barge-in with an explanation of the violation.
Multi-Interface Navigation: Expanding beyond Jira and Google Earth to visually navigate additional production tools commonly found in telecommunications NOC environments—ticketing systems, provisioning platforms, and network management dashboards—using the same vision-first approach with no API dependencies.

Phase 2: Enterprise Execution Coverage

Extending Circuit Stitcher's execution capabilities across all physical network domains:

Longhaul underground fiber
Overground routes
Subsea marine cable systems

This phase integrates directly into telecommunications NOC execution workflows, ensuring that Circuit Stitcher can physically reroute circuits across every cable type and operational context an enterprise network team encounters.

Circuit Stitcher ingests structured execution payloads from the upstream Aegius nodes and adapts its geospatial reasoning to the specific physical constraints of each environment—mountain passes for overground, seabed topology for subsea, and conduit sharing for underground.

The goal: No reroute instruction, on any cable type, in any region, is executed without Circuit Stitcher validating the physical path against SRLG policy, terrain constraints, and latency budgets—eliminating the Execution Gap across the entire infrastructure portfolio.

Phase 3: Planetary-Scale Execution

Standards Alignment & Industry Vision

Anchoring to ITU ION-2030

The North Star of the Aegius Framework is the ITU ION-2030 standard (GSTR.ION-2030)—International Optical Networks towards 2030 and Beyond—developed by ITU-T Study Group 15, the expert group responsible for standards on networks, technologies, and infrastructures for transport, access, and home.

At its core, the framework emphasizes the mutual empowerment of AI and optical networking:

AI enhancing optical networks: Using digital twins, multimodal learning, and autonomous control to enhance network reliability, reduce energy use, and anticipate service needs.
Optical networks enabling AI: Providing the high-capacity, low-latency, and deterministic connectivity needed for distributed AI training, real-time inference, and data exchange between cloud and edge.

ION-2030's Four Principal Advances

The framework envisions a service-oriented architecture combining high performance with intelligence and sustainability. It highlights four principal advances—each of which the Aegius Framework is designed to support:

Terabit-per-second connectivity and sub-millisecond latency to support emerging digital services.
Integrated sensing, computing, and AI agents within optical layers for real-time awareness and automation.
Energy-efficient and quantum-resilient designs to ensure long-term security and sustainability.
End-to-end service optimization across multiple network domains and layers.

Circuit Stitcher aligns directly with the second and fourth advances. Integrated AI agents within optical layers are only effective if they can translate awareness into physical action. End-to-end service optimization across multiple domains requires an execution layer that can physically reroute infrastructure—not just recommend a change, but perform it within policy boundaries and with full auditability.

Circuit Stitcher is the autonomous execution layer that closes the gap between ION-2030's intelligent sensing vision and the physical reality of production network infrastructure.

📎 Reference: ITU ION-2030 — Enabling the Intelligent Optical Network Era

Potential Impact: Extreme Resilience

At full maturity, the Aegius Framework could support deployment in extreme reliability environments—such as Japan's seismic-active subsea cable corridors. Executing an autonomous reroute within seconds of a seismic event—validated against SRLG policy, terrain constraints, and latency budgets—while simultaneously generating an executive-ready briefing through SIB, represents a long-term north star for the framework's execution capabilities.

Phase 4: Aegius Planetary Governance

Risk Acknowledgment & Orchestration Governance

The Cost of a Wrong Execution

In a telecommunications environment, the cost of a wrong decision is measured in dropped calls, lost revenue, and regulatory exposure. The "hallucination risk" inherent in any AI system is most dangerous at the execution layer:

A diagnostic error in the sensing layer is dangerous.
A translation error in the briefing layer is catastrophic.
An execution error at the infrastructure layer is irreversible—because it physically alters production network state, potentially creating the very Shared Risk Link Groups it was designed to prevent.

Circuit Stitcher occupies the most consequential position in the Aegius intelligence chain. Acknowledging this risk head-on is not a weakness—it is the foundation of the governance model.

Policy Firewall Architecture

The Aegius Framework is not a generic automation tool—it is a strict Policy Firewall for AI agents.

This phase formalizes the orchestration governance layer that ensures every autonomous action is auditable, bounded, and grounded in operator-defined policy.

Key governance principles:

Human-in-the-Loop: Critical circuit changes require human verification before execution, ensuring AI augments human judgment rather than replacing it. Circuit Stitcher's Absolute Override Protocol is the enforcement mechanism—it blocks unauthorized actions at the point of execution, not after the fact.
Auditability: Every mouse movement, polyline stitch, and reroute decision is logged and traceable to operator-defined policy. Circuit Stitcher generates a complete execution audit trail that feeds back into SIB for executive-readable compliance reporting.
Bounded Autonomy: Agents operate within strict policy boundaries—no unbounded exploration of production network state. The Hierarchy of Truth ensures that RTT telemetry and SRLG policies override any manual input that contradicts the physics of the fiber route.

Benchmarking & Accountability

Self-Imposed SLA Targets

Governance principles are only credible when they are backed by measurable commitments. Rather than asking stakeholders to evaluate the Aegius Framework on claims alone, the following SLA targets define the quantified contract against which the framework should be held accountable.

These targets are derived from the core Beyond the Topology Illusion research and grounded in inference economics benchmarks and planetary-scale network standards including ION-2030.

Service Category	SLA Target	Metric Definition	Strategic Impact
System Availability	99.999%	"Five 9s" uptime for the Aegius Control Plane and sensing nodes.	Ensures zero-gap protection for infrastructure costing $14,056/min during downtime.
Interaction Velocity	< 500ms	Real-time audio/vision response delay from Gemini 2.5 Flash nodes.	Prevents the 30% engagement erosion caused by high-latency AI interactions.
Audit Velocity	< 10 Minutes	Time to complete a planetary-scale circuit audit (Sense → Translate → Stitch).	Represents a 99% reduction from the traditional 2–3 week manual audit cycle.
Diagnostic Precision	Sub-ms Error Delta	Alignment with ION-2030 standards for deterministic latency detection.	Resolves the "Topology Illusion" by ensuring RTT values match physical reality.
Financial Safeguard	< $4,800/mo	Maximum allowable exposure to the Recompute Tax per GPU cluster.	Directly protects the Inference Moat and operational profitability.
Briefing Delivery	< 10 Seconds	Time to generate an interleaved multimodal storyboard for executives.	Collapses the "Communication Void" between the NOC and the boardroom.

In addition, Vertex AI Grounding integrates Google Search to provide real-time competitive analysis of global token pricing and regional SLA benchmarks across 111 countries.

📎 Full benchmarking document: Aegius Benchmarking: Agreed Sample SLAs

AI Sovereignty & Geopolitical Context

Geopolitical Awareness & Regulatory Foresight

Telecommunications is the new front line of geopolitics.

In 2026, organizations face mounting pressure around:

Data residency
Foreign control of critical infrastructure
The regulatory landscape exemplified by frameworks like the EU AI Act

The question of AI Sovereignty—who owns the reasoning, who controls the model, and where the data lives—remains fluid and murky.

The Aegius Framework is designed with this reality in mind. By enabling operators to run inference on-premises, the framework ensures that both the reasoning and the data remain within national borders. For Circuit Stitcher specifically, this means that every execution decision—every reroute, every override, every polyline stitch—is made within sovereign infrastructure, with no execution telemetry leaving national boundaries.

However, this is not a solved problem. True AI Sovereignty must be approached carefully, in partnership with different geographical regions and nations, navigating evolving regulatory requirements over time.

The Aegius Framework provides the architectural foundation for that journey—a governance-first design that adapts as the geopolitical landscape shifts.

The Financial Case

Total Cost of Ownership & Revenue Protection

Metric	Traditional Ops	Aegius Framework	Financial Impact
Downtime Cost	$33,333 / minute	80% Reduction	$26,667 saved / min
Execution Time	2–4 Hours (Manual Reroute)	< 10 Minutes (Autonomous)	95% Time Reduction
SRLG Incidents	Undetected Until Failure	Blocked at Execution	Prevents cascading outages
GPU Utilization	75% (Jitter/Latency)	93% (Optimized)	$32,400 saved / month / cluster
MTTR	Hours/Days	Minutes	40% Efficiency Gain

At scale, these savings compound dramatically:

Telecommunications downtime costs an average of $2 million per hour, and enterprise-class outages can reach $5 million per hour excluding fines or penalties.
Every hour spent on manual rerouting is an hour of unmitigated downtime exposure, compounding SLA penalties, and accumulating the Recompute Tax across the GPU grid.
A single undetected SRLG—the kind Circuit Stitcher's Override Protocol is designed to prevent—can turn a localized fiber cut into a catastrophic dual-path failure affecting millions of active sessions.

The Aegius Framework—with Circuit Stitcher as its autonomous execution layer—represents millions in annual savings from accelerated rerouting, SRLG prevention, reduced MTTR, and audit-ready compliance reporting.

This positions the framework not as a cost center, but as infrastructure-level revenue protection.

Path to Productization

The path to productization follows two complementary models:

Usage-Based Pricing: Per reroute executed, per SRLG conflict blocked, or per circuit stitched—aligning cost directly with value delivered. This model lowers the barrier to entry for operators who want to validate ROI before committing.
Feature-Based Tiers: A tiered structure that scales with operational maturity—starting with Latency Lens for diagnostic sensing, adding Situation Intelligence Brief for executive translation and compliance-ready reporting, and unlocking the full Aegius Framework with Circuit Stitcher and Policy Firewall governance for end-to-end autonomous resilience.

Call to Action

Adopt Circuit Stitcher, and close the Execution Gap between diagnostic intelligence and physical infrastructure action.

Stop letting manual context-switching create hidden Shared Risk Link Groups—start executing policy-bounded, autonomous reroutes in minutes, powered by Google Cloud, and let an AI co-pilot protect your network at the point where decisions become physical reality.

Note: All technical metrics and financial benchmarks cited in this submission are derived from the following research documents and established industry benchmarks:

Beyond the Topology Illusion

Evaluating AI Agent Governance Idea

AI for Network Resilience Assessment

Established industry benchmarks for inference economics

Built With

gemini-2.5-flash-native-audio
gemini-live-api
google-cloud-run
google-earth
imagecaptture-api
imagen
medium
nano-banana-2
python
v1alpha
vanilla-html
web-audio-api
websocket