AiOS→HR / People→Operate & Support→HR09

Employee Relations Case Summary & Risks

Summarize employee-relations cases, cite facts against policies, and surface compliance risks and cross-case patterns.

Consistent case documentation reveals recurring patterns and compliance gaps earlier, reducing escalation risk and giving leadership a clearer basis for policy decisions.

GenAI Impact

50%

Faster

7.5

Hours saved

15.1

Hours without AI

Based on: 5 employee relations cases reviewed against compliance framework

Enforces verbatim citation of source case files and policy-anchored severity classifications across all cases, ensuring every risk assessment traces to documented evidence rather than inconsistent analyst interpretation.

Prevents protected employee relations data — disciplinary records, grievance details, and performance improvement plans — from exposure in unapproved GenAI tools through mandatory redaction gates and approved-platform restrictions.

Before You Start

This workflow processes employee relations case files containing protected personal data (disciplinary records, grievance details, performance improvement plans). Do not paste these inputs into public or unapproved GenAI tools.

GenAI may conflate facts from separate cases or misclassify policy relevance. Verify every AI-generated case summary and risk highlight against the original case files and organisational policies before distribution.

Who's Involved

HR Specialist

Runs the case review workflow and produces the case trend summary

Employee Relations Manager

Reviews risk classifications and approves escalation decisions

Legal Advisor

Reviews compliance-related risks and confirms regulatory classification

Execution Steps

HumanGenAIHybrid

Before you start

Confirm all employee relations case files for the review period are collected and accessible
Confirm engagement risk flags from the prior survey analysis are available
Confirm your organisation's policy and compliance framework is current and approved
Confirm escalation and severity criteria are documented and agreed upon
Confirm you are using an approved GenAI tool for this workflow

Inputs

Employee Relations Case Filesdownload
Engagement Risk Flagsdownload
From HR17 — Engagement Survey Insights & Actions
Organisation Policy & Compliance FrameworkdownloadEscalation & Severity Criteriadownload

Prompt

Define structured template for case summarisation

CONTEXT
You will be provided with the following source documents:
1. Employee Relations Case Files
2. Engagement Risk Flags
3. Organisation Policy & Compliance Framework
4. Escalation & Severity Criteria

TASK
Define a structured markdown template for summarising individual employee relations cases. The template must capture case type, timeline, key parties by role title, documented facts with source references, relevant policy sections, and current case status.

OUTPUT FORMAT
Produce a template in this exact structure:

## [Case Reference]
### Case Overview
- **Case Type:** [Grievance / Disciplinary / Performance / Workplace Conflict / Other]
- **Date Opened:** [Date]
- **Current Status:** [Open / Under Review / Escalated / Closed]
- **Parties Involved:** [Role titles only]

### Documented Facts
| # | Date | Event | Source Document | Verbatim Excerpt |
|---|---|---|---|---|

### Relevant Policies
| Policy Section | Relevance to Case |
|---|---|

### Open Items
- [List unresolved actions or pending decisions]

CONSTRAINTS
Do not add fields beyond those specified. Do not include any personally identifiable information in the template example. Do not create scoring or automated decision fields.

Outputs

Case Summary Template
AI-drafted · you verify·passed to next step

Verification: Confirm the template fields align with your organisation's case documentation standards before proceeding

Before you start

Confirm sensitive personal identifiers have been redacted from case files before prompting

Data Handling: Do not paste case files containing personal identifiers such as home addresses or medical details into the prompt; redact these fields first

Inputs

Employee Relations Case Filesdownload
Case Summary Templatefrom prev step

Prompt

Summarise each case with verbatim evidence

CONTEXT
You will be provided with employee relations case files and a case summary template defining how each case should be documented.

TASK
For each case file, populate the case summary template. Extract every documented fact with its date and source reference. Quote relevant passages verbatim from the case file. Identify all policies referenced or implicated.

OUTPUT FORMAT
For each case, produce a completed template following the exact structure provided. Every entry in the Documented Facts table must include the source document name and a verbatim excerpt. List all policies referenced or implicated in the Relevant Policies table.

CONSTRAINTS
Do not summarise or interpret facts — extract text exactly as written in the source case file. Do not add risk assessments at this stage. Do not include names of individuals; use role titles throughout.

Outputs

Individual Case Summaries
AI-generated·passed to next step

Verification: Verify every documented fact exists verbatim in the source case file before proceeding to analysis

Inputs

Individual Case Summariesfrom prev step
Organisation Policy & Compliance Frameworkdownload

Prompt

Prompt available with library accessGet Access →

Outputs

Cited Evidence Sheets
AI-generated·passed to next step
Every fact from the individual case summaries appears in the policy-mapped or unmapped section
No fact is classified as a potential violation without a specific policy section reference
Unmapped facts are listed with clear reasoning for the lack of mapping

Verification: Verify the AI has not fabricated policy section references — confirm every cited policy section exists in your framework

Inputs

Cited Evidence Sheetsfrom prev step
Engagement Risk Flagsdownload
From HR17 — Engagement Survey Insights & Actions

Prompt

Prompt available with library accessGet Access →

Outputs

Case-Engagement Cross-Reference
AI-generated·passed to next step
Every engagement risk flag from the input appears in the table or under Unmatched Flags
Correlation strength ratings are supported by cited case evidence

Inputs

Cited Evidence Sheetsfrom prev step
Escalation & Severity Criteriadownload

Prompt

Prompt available with library accessGet Access →

Outputs

Risk Highlight Report
AI-generated·passed to next step
Every risk cites the specific escalation criterion or policy reference used
Severity levels match the escalation and severity criteria definitions
No employment action recommendations appear — risks and evidence only

Verification: Verify the AI has not assigned severity levels inconsistently across similar findings in different cases

Inputs

Risk Highlight Reportfrom prev step
Case-Engagement Cross-Referencefrom prev step

Prompt

Prompt available with library accessGet Access →

Outputs

Pattern & Escalation Analysis
AI-drafted · you verify·passed to next step
Every pattern is supported by evidence from at least two cases or one critical-severity case
Escalation triggers reference specific case evidence and monitoring actions
No individual names or personally identifiable details appear in the pattern analysis

Inputs

Pattern & Escalation Analysisfrom prev step
Individual Case Summariesfrom prev step

Prompt

Prompt available with library accessGet Access →

Outputs

Case Trend Summary
AI-drafted · you verify·passed to next step
Executive overview accurately reflects the aggregate patterns and risk posture
Case volume table accounts for every case in the confirmed review scope
Every pattern cites specific case evidence and policy implications
Next-step options are tied to pattern analysis with clear rationale

Verification: Verify the summary does not contain disciplinary recommendations — it must present trends, risks, and options only

Reference

Guardrails

  • Factual Citation Only — Only cite documented facts from official case files — never let the AI infer undocumented events or assume intent from ambiguous notes.
  • Separate Fact From Interpretation — Keep factual case summaries strictly distinct from risk assessments and recommendations to prevent conflation in downstream decision-making.
  • Policy-Anchored Assessment — Every risk classification must reference a specific policy section or escalation criterion, not general judgment or subjective severity.

Pitfalls

  • Pasting employee relations case files containing protected personal information into a non-approved GenAI tool
  • Treating AI-generated risk classifications as final compliance determinations without HR or legal review of the source findings
  • Allowing the AI to infer undocumented events or assume intent from ambiguous case notes and incomplete records
  • Merging findings from separate cases into a single summary, obscuring individual case boundaries and accountability trails

Definition of Done

  • Every case summary cites the specific source case file and documents facts with verbatim excerpts
  • The Case Trend Summary covers all cases from the confirmed review scope with no omissions
  • Every risk highlight references a specific policy section or escalation criterion from the organisation's framework
  • No hallucinated facts, dates, or policy references appear in any AI-generated output

Unlock the Full Library

Get full access to all prompts, execution steps, and downloadable examples — for this playbook and the rest of our GenAI capability framework — AGASI AiOS.

We'll send a magic link — no password needed.

AGASI AiOS · HR09 v1.0 · Apr 7, 2026