Skip to content
DA DataAcuity by The Geek Network

DataAcuity — Circle Product Family Integration

Aether by Circle | CircleOS | CircleOne | CircleUp

Classification: Internal Links: AppInfo/DataAcuity/DataAcuity_Internal_Analytics.md — analytics framework Links: AppInfo/DataAcuity/DataAcuity_Data_Flow.md — TGN app event schemas Links: .claude-memory/circleone-keyboard.md — CircleOne keyboard notes Links: .claude-memory/aether-mesh.md, .claude-memory/aether-node-tipping.md Links: AppInfo/CircleUp/CircleUp_Product_Specification.md — CircleUp definition (complete) Links: AppInfo/CircleUp/CircleUp_Business_Operations_Skills_Map.md — full skills map per department


Why the Circle Products Need DataAcuity Integration

The Circle product family is the infrastructure layer of the entire ecosystem:

  • Aether by Circle = the network (how everything connects offline)
  • CircleOS = the platform (where apps run for power users/developers)
  • CircleOne = the input (how people type in their language)
  • CircleUp = the knowledge platform + agent runtime (how the organisation runs autonomously)

Without visibility into how these products perform and how users engage with them, we cannot make good decisions about the infrastructure that underpins the whole mission.

DataAcuity's internal analytics layer must cover these products with the same rigour as the consumer apps. The Circle products are not optional add-ons — they are the substrate.


Aether by Circle — Network Analytics

What Aether is (for analytics context)

Aether by Circle is the mesh networking layer. Every TGN app uses it. When users have no internet, Aether provides connectivity via:

  • BLE (Bluetooth Low Energy) for short-range
  • Wi-Fi Direct for medium-range
  • LoRa (future) for long-range

Relay nodes are volunteers who enable others to route traffic through them. They are tipped in Qi (settled to SDPKT wallet via LedgerAPI).

Why Aether analytics matter for DataAcuity (external)

The Aether mesh creates a real-time map of where South Africa goes offline. This is one of DataAcuity's most unique signals — no other provider has this. See DataAcuity_Full_Proposition.md — "The Aether Advantage" section.

Why Aether analytics matter internally

If Aether is failing (low delivery rates, high latency, sparse nodes), the entire ecosystem degrades for our most vulnerable users — those without data. We must know this before they do.

Aether event schema (internal analytics)

// AetherNetworkEvent
{
  "eventType": "AetherNetworkEvent",
  "timestamp": "2026-04-18T14:22:00Z",
  "nodeId": "<node-pseudonymous-id>",
  "eventSubtype":
    "node-online |
     node-offline |
     message-sent |
     message-delivered |
     message-failed |
     relay-volunteered |
     relay-session-started |
     relay-session-ended |
     relay-tip-paid |
     relay-tip-failed |
     coverage-gap-detected |
     fallback-to-internet",
  "protocol": "ble | wifi-direct | lora | hybrid",
  "hopCount": 2,
  "latencyMs": 340,
  "payloadSizeBand": "0-1KB | 1-10KB | 10-100KB | 100KB+",
  "deliverySuccess": true,
  "locationZone": "GP-JHB-SOUTH",
  "appContext": "txtMe | SDPKT | TagMe | Panik | SleptOn | other",
  "relayTipAmountQi": 5
}

Aether KPIs

KPI Target Alert threshold Business value (external DataAcuity)
Active relay nodes Growing MoM Declining 2+ months Coverage map completeness
Message delivery rate (mesh) > 85% < 70% Signals reliability
Median latency (1 hop) < 500ms > 2000ms User experience proxy
Coverage zones (active nodes) Growing New zero-coverage zones Infrastructure gap map
Relay tip payment success > 97% < 94% Volunteer incentive health
Internet fallback rate < 40% > 60% (mesh not functioning) True offline capability
New volunteer nodes/month Growing Declining Community participation

Aether coverage gap map (external DataAcuity product)

Real-time aggregation of:

  • Zones with zero active relay nodes (no mesh coverage)
  • Zones with high internet-fallback rates (mesh present but insufficient)
  • Zones with high offline-only activity (Aether carrying the full load)

This map is sold to telecoms, government, and development finance as infrastructure insight. It is the only real-time network coverage gap map built from ground-truth user behaviour.


CircleOS — Developer Platform Analytics

What CircleOS is (for analytics context)

CircleOS is the operating layer — a platform for power users and developers who want deeper access to the TGN ecosystem. Developers build on CircleOS. Advanced users run it. It is the API and runtime layer that other Circle products build on.

Why CircleOS analytics matter

CircleOS adoption = developer ecosystem health. If developers aren't building on CircleOS, the ecosystem doesn't compound. We need to know: who is building, what they're building, where they're hitting friction.

CircleOS event schema

// CircleOSEvent (internal analytics)
{
  "eventType": "CircleOSEvent",
  "timestamp": "2026-04-18T14:22:00Z",
  "sessionId": "<session-pseudonymous-id>",
  "eventSubtype":
    "device-registered |
     developer-api-call |
     app-installed |
     app-launched |
     app-uninstalled |
     update-available |
     update-installed |
     update-declined |
     api-error |
     permission-granted |
     permission-denied |
     mesh-transport-used",
  "osVersion": "1.2.3",
  "apiEndpoint": "/api/mesh/send | /api/wallet/balance | /api/identity/verify | ...",
  "apiResponseCode": 200,
  "apiLatencyMs": 45,
  "appCategory": "finance | messaging | safety | productivity | entertainment | other",
  "locationZone": "GP-JHB-SOUTH",
  "source": "internet | aether"
}

CircleOS KPIs

KPI Target Alert threshold
Registered developer devices Growing MoM Flat or declining
API call success rate > 99% < 97%
API p95 latency < 200ms > 500ms
Update adoption rate (within 48h) > 60% < 30%
Apps installed per device (median) > 3 < 2
Most-called API endpoints Monitor No change (signals stagnation)
API error rate by endpoint < 1% per endpoint > 3% (endpoint-specific)

Developer friction signals

  • High error rate on specific API = undocumented behaviour or breaking change
  • Low app install rate = discoverability problem on CircleOS app store
  • High update-declined rate = update is too large / breaking / untrusted
  • Low developer retention = onboarding friction or insufficient documentation

CircleOne — Keyboard Analytics

What CircleOne is

CircleOne is the keyboard app. Cross-platform. Built with .NET MAUI. Key feature: isiBheqe Unicode input (supporting click consonants and other characters absent from standard keyboards).

Separate repo — see .claude-memory/circleone-keyboard.md for build details.

Why CircleOne analytics matter

CircleOne is infrastructure for language inclusion. If the keyboard is buggy, slow, or doesn't support the characters users need, the entire multilingual promise of The Geek Network falls apart at the input layer.

Specifically: if isiZulu, isiXhosa, or isiBheqe users can't type comfortably, they default to English — and the localisation investment across all 17 apps is undermined.

CircleOne event schema

// CircleOneEvent (internal analytics)
{
  "eventType": "CircleOneEvent",
  "timestamp": "2026-04-18T14:22:00Z",
  "sessionId": "<session-pseudonymous-id>",
  "eventSubtype":
    "keyboard-opened |
     keyboard-dismissed |
     language-switched |
     layout-switched |
     custom-key-used |
     isibheqe-character-used |
     autocorrect-accepted |
     autocorrect-rejected |
     word-deleted-after-entry |
     emoji-used |
     clipboard-paste",
  "fromLanguage": "en",
  "toLanguage": "zu",
  "inputMethod": "tap | swipe | voice",
  "errorRate": 0.08,
  "sessionDurationSeconds": 240,
  "appContext": "txtMe | Bruh | other",
  "osType": "android | ios",
  "deviceCategory": "low-end | mid | high"
}

CircleOne KPIs

KPI Target Alert threshold
isiBheqe character usage rate Growing Flat (Unicode not working)
Language switch events per session Monitor Sudden drop (lang support problem)
Autocorrect rejection rate < 20% > 40% (autocorrect is wrong too often)
Word-deleted-after-entry rate < 15% > 25% (input accuracy problem)
Keyboard session duration Growing Declining (users switching to other keyboard)
Low-end device error rate < 5% > 10% (performance on cheap devices)
isiZulu / isiXhosa switch rate Growing Declining (language adoption regression)

isiBheqe adoption tracking

isiBheqe (the emerging Unicode standard for click consonants) is a core differentiator. We need:

  • Daily active isiBheqe users (DAIBU)
  • Most-used isiBheqe characters (which clicks are needed most)
  • Input method for isiBheqe (long-press? custom key? swipe?)
  • Error rate for isiBheqe characters (are they rendering correctly downstream?)

Reference: AppInfo/CircleOne/isibheqe-keyboard-project-plan.docx


CircleUp — Agent-Native Knowledge Platform Analytics

What CircleUp is (for analytics context)

CircleUp is the open-source, agent-native knowledge platform that runs The Geek Network as an autonomous organisation. It is "The Organisation" vault — staffed by AI department agents (Engineering, Operations, Product, Marketing, Finance, Support, QA), orchestrated through Markdown files on disk.

Two products in one:

  1. Internal runtime — The Geek Network runs on CircleUp. Every business function automated by a department agent.
  2. Open-source product — available to solopreneurs and small teams. MIT license. Blazor Hybrid (Android + iOS + Web).

See AppInfo/CircleUp/CircleUp_Product_Specification.md — full product definition. See AppInfo/CircleUp/CircleUp_Business_Operations_Skills_Map.md — full skills map per department.

Why CircleUp analytics matter

CircleUp is both the brain of the business AND a product. Internal analytics serve two purposes:

  1. Self-referential — CircleUp analytics tell us if agents are working, skills are effective, and knowledge is growing
  2. Product validation — before publishing skills to B! via SleptOn, we need proof they work

CircleUp event schema

// CircleUpEvent (internal analytics)
{
  "eventType": "CircleUpEvent",
  "timestamp": "2026-04-18T14:22:00Z",
  "sessionId": "<session-pseudonymous-id>",
  "eventSubtype":
    "vault-opened |
     note-created |
     note-updated |
     note-deleted |
     link-created |
     link-broken |
     search-performed |
     search-zero-results |
     skill-invoked |
     skill-succeeded |
     skill-failed |
     agent-task-created |
     agent-task-completed |
     agent-task-escalated |
     graph-viewed |
     sync-completed |
     skill-published-to-slepton",
  "department": "engineering | operations | product | marketing | finance | support | qa | none",
  "skillName": "deployment/iis-deploy | coding/review | finance/reconciliation | null",
  "agentModel": "claude-opus | claude-sonnet | butler | local | null",
  "taskResolutionType": "autonomous | human-assisted | escalated | null",
  "noteCount": 347,
  "skillCount": 48,
  "graphNodeCount": 347,
  "graphEdgeCount": 892,
  "platform": "web | android | ios | windows | macos"
}

CircleUp KPIs (internal)

KPI Target Alert threshold
Notes created per week (team) Growing Flat > 4 weeks (knowledge capture stopping)
Agent task completion rate (autonomous) > 80% < 60% (agents not effective)
Skill invocation success rate > 90% < 75% (skill quality problem)
Human escalation rate < 20% > 40% (agents overwhelmed or under-skilled)
Graph density (edges/nodes ratio) > 2.5 < 1.5 (notes not being linked)
Search zero-result rate < 15% > 30% (knowledge gaps)
Skills added per month Growing Declining (curation slowing)
Skills graduated to B! (cumulative) Growing Stalled (curation pipeline not producing)

External DataAcuity signals (future)

CircleUp is internal-only for now. In Phase 4 (B2B product), multi-tenant CircleUp deployments will generate anonymised knowledge graph health signals — graph density, skill adoption rates, agent effectiveness by industry vertical — for DataAcuity's enterprise analytics layer.

These signals will be unique: no other product captures how organisations build and maintain their institutional knowledge at this level of granularity.


Across All Circle Products — The Cross-Product Journey

The most powerful analytics signal is not per-product — it's the cross-product journey.

Example journeys to track:

  • CircleOne (keyboard) → txtMe! (uses keyboard to message) → Aether (sends via mesh)
  • CircleOS (developer builds app) → SleptOn (distributes app on SleptOn) → Aether (users download via mesh)
  • Aether (mesh active) → SDPKT (payment via mesh) → Qi (earned for offline transaction)

Cross-product signals reveal:

  • Which Circle products drive TGN consumer app adoption
  • Whether improving CircleOne increases txtMe! session depth
  • Whether Aether coverage growth correlates with SDPKT transaction growth in offline zones

These cross-product insights feed BOTH internal product decisions AND the external DataAcuity "ecosystem-level signals" that enterprise clients cannot get anywhere else.

Something went wrong on this page. Reload