SAFEATWORK carries employee concerns through a five-step structured lifecycle — Intake to Closure — with CLEARS profiling and a human-in-the-loop AI layer that sees what isolated tickets miss.
5
Structured steps
9d
Avg cycle time
61%
AI drafts accepted
SAW-ER-2031Active case
Step 4 of 5 · Investigation
Structured inquiry underway
Investigator working from fixed interview template. CLEARS R-score flagged at 44/100 — 2 prior flags on respondent already on file.
📋 Template active⚠ R-score 44/100👥 2 witnesses
How it works
One platform. Five steps. Zero blind spots.
Built for ER managers who need structure, speed, and a defensible record — from the first complaint to the final closure report.
📥
Structured Intake
Complainant submits their version through a guided form. ER manager reviews and logs — never opens a case blind.
📊
Auto CLEARS Report
As soon as intake is complete, CLEARS profiles auto-surface for both parties. Assessment begins from a pattern, not a blank page.
🤖
Human-in-Loop AI
AI drafts summaries and flags patterns across cases. Every output pauses for human review before becoming a record.
Interactive Walkthrough
Live case — SAW-ER-2031
Click each step or use the buttons inside each panel. All actions are live — forms save to state, CLEARS reports render dynamically, AI review responds to your choices.
CASE WORKFLOW · SAFEATWORKSAW-ER-2031
1
Intake & Documentation
2
Triage & Assessment
3
CLEARS
4
Investigation
5
Closure
Loading…
The Framework
CLEARS — Six Signals, Structural Intelligence
CLEARS is not a survey. It is a continuous behavioural readout built from LMS data, attendance, peer reviews, and delivery metrics — surfaced the moment a case opens.
In a live case: CLEARS profiles auto-generate for both the complainant and respondent at Step 3 (CLEARS). The investigator sees trend lines — not just current scores — before a single interview is conducted.
🔐 DPDP Act 2023 compliance: CLEARS scores constitute inferred personal data under India's Digital Personal Data Protection Act. Processing requires a lawful basis (employment purpose or consent). Employees have the right to access, correct, and request erasure of their CLEARS data where retention is not required by law. AI-generated summaries in Step 4 (Investigation) must not solely determine employment outcomes — SAFEATWORK's human-in-the-loop architecture is designed to satisfy this requirement.
CLEARS Intelligence Report
Auto-generated at case intake · Read-only until closure
SAW-CLR-2031-A
👩
Priya Mehta
Senior Analyst · Finance · Bengaluru · Joined 2019
Complainant
Legal Reference
Employment Regulations Reference
Key statutes and frameworks that govern employee relations across India, EU, and USA. SAFEATWORK maps each case type to applicable law automatically.
India's employment relations framework is plural — layered across central and state legislation. Key statutes relevant to workplace ER are listed below. SAFEATWORK flags applicable acts at intake based on case type and employee category.
The POSH Act 2013 (Prevention of Sexual Harassment) mandates Internal Committees (IC) in organisations with 10+ employees. SAFEATWORK manages POSH cases in a separate, confidential workflow with IC-specific timelines.
EU employment law operates at two levels: EU Directives (which Member States implement into national law) and directly applicable Regulations. Below are the four key Directives relevant to ER in EU-based operations.
US employment law is federal and state. Federal statutes enforced by the EEOC cover discrimination, harassment, and protected classes. State laws may extend protections further — SAFEATWORK flags state overlaps where configured.
Closure Intelligence
What every closed case feeds back
Closure doesn't archive a file — it updates org-wide CLEARS scores, feeds operational metrics, and surfaces hotspots before a fourth case arrives.
Cases closed this month
14
↑ 3 vs last month
Avg. cycle time (days)
9d
↓ 2d vs target
Teams flagged on R score
2
⚠ Escalation advised
AI drafts accepted as-is
61%
↑ 8% vs Q1
Open Cases
CLEARS Hotspots — Relationship Score
Submit Your Concern
Your submission is confidential. An ER manager will review it within 2 working days.
ER Manager — Triage & Assessment
Review the complainant's submission, assign case details, and trigger the CLEARS reports.
Respondent Notification
The respondent is formally notified of the case and invited to provide their account. CLEARS report is auto-attached (read-only).
⚠️
Notification to: Rajesh Kumar (Manager · Finance)
You are named as a respondent in case SAW-ER-2031. This is a confidential process. You are required to submit your account within 5 working days. The investigation is neutral and no outcome has been determined.
Investigation Workbench
Structured, neutral interview template. Same questions every case — ensuring consistency and legal defensibility.
CLEARS Report — Respondent
Rajesh Kumar · Manager · Finance · Auto-generated at intake
AI Pattern Review — Pending Human Sign-off
Case SAW-ER-2031 · Step 4 of 5 · Investigation — Auto-generated draft, no record until approved
AI DraftSAW-ER-2031 · Generated just now
Pattern detected: This is the 3rd case involving Manager Rajesh Kumar in 8 months. Relationship (R) CLEARS score has declined from 67 → 44 over that period (−23 pts). The complainant's Engagement (E) score is also declining (71 → 62), consistent with a hostile environment pattern.
Investigation finding: Preliminary evidence is partially substantiated. Two witnesses corroborate the complainant's account on the September incident. Respondent's account contradicts on key material facts.
Recommendation: Escalate to a team-level culture review. Consider interim protective measures for complainant. HR Business Partner for the Finance division should be briefed before closure.
⚠ This draft is not a record. It becomes binding only when an authorised ER manager clicks Accept.
Case Closure — SAW-ER-2031
Closing this case will update CLEARS scores, generate the closure report, and notify both parties.