Capture
Collect delivered, opened, clicked, submitted, and reported events.
Evidence-led human risk · TaiGong 1.3.4
Move from click rates to a human-risk operating loop.
01 / Risk model
Observe · Prioritize · Follow up
Human risk starts with explainable evidence.
A click can come from an employee, a mail gateway, a link scanner, or prefetch. TaiGong preserves the event sequence and adds licensed context before the team decides what action to take.
Contextualize
Review recipient timelines and available browser, device, network, and location evidence.
Prioritize
Use core risk signals and licensed advanced views to identify follow-up needs.
Train
Place selected users into licensed awareness courses, batches, and assessments.
Review
Use later campaigns, exports, and automation history to evaluate operational change.
02 / Delivered boundaries
Core · Licensed · Integration-only
Risk is a workflow, not one unlocked screen.
Version 1.3.4 separates the evidence and risk baseline from licensed analytics and follow-up modules.
| Human-risk capability | Status | Delivered in 1.3.4 |
|---|---|---|
| Campaign events and recipient timelines | Core | Delivery, open, click, submission, report, and result evidence. |
| Basic risk signals and campaign evidence | Core | Current campaign and recipient evidence used for risk review. |
| Advanced dashboard, department exposure, trends, rankings, and higher-risk lists | Licensed | Available with the advanced analytics entitlement. |
| Awareness courses, training batches, assessments, and completion views | Licensed | Available with the awareness/scenarios entitlement. |
| Stages, scheduled rules, event triggers, and execution history | Licensed | Available with the automation entitlement. |
| PDF reporting and advanced exports | Licensed | Available where the reporting capability is enabled. |
| Slack notifications | Integration-only | Slack can receive selected platform events; it is not a simulation-delivery channel. |
| Slack or Microsoft Teams exercise delivery | Unsupported | Not delivered in 1.3.4. |
| Autonomous AI risk agent | Unsupported | Current AI capability is licensed assistance, not an autonomous agent. |
03 / Measurement quality
Do not reward noisy metrics
Separate employee behavior from automated traffic.
Security gateways, link scanners, cloud systems, and prefetch can distort campaign metrics. Where licensed, TaiGong environment and quality controls provide additional evidence for reviewing those events.
Keep the sequence visible.
Review when a message was delivered, opened, clicked, submitted, or reported, together with the recipient result and available evidence.
Add context before classification.
Licensed environment controls can use gateway, prefetch, cloud, device, network, JA4, and quality signals to support review.
Prioritize without treating one event as a verdict.
Use core evidence and scores, then add licensed advanced views to review behavior across users, departments, and campaigns.
Prepare management and audit outputs.
Licensed analytics and reporting support summaries, PDF or CSV outputs, and campaign comparisons. The platform does not claim that one score proves individual intent.
04 / Targeted follow-up
Licensed awareness and automation
Give the next action a clear owner.
The licensed awareness module manages courses, training batches, assessments, and completion views. The licensed automation module adds scheduled and event-based actions with reviewable execution history.
From evidence to action
Build a risk program around facts.
Review the core evidence model first, then decide which analytics, training, automation, and reporting modules your program needs.