HRFLOW — HR Performance Intelligence

Total Agents

6

3 extraction · 3 decision

Tests Passing

12 / 12

100% pass rate

Reviews Analyzed

24

Q3 review cycle

Promotion Ready

4

Across all teams

Agent Confidence Scores

ReviewParser

—

GoalExtractor

—

FeedbackAnalyzer

—

TaskPrioritizer

—

OwnerAssigner

—

EscalationDecider

—

Recent Reviews

Employee	Role	Rating	Action
Sarah Martinez	Senior Engineer	Exceeds	Promotion
James Liu	Staff Engineer	Meets	On Track
Priya Sharma	Engineering Mgr	Exceeds	Promotion
Alex Thompson	Junior Engineer	Needs Imp.	PIP

Goal Tracking Summary

Employee	Goal	Progress	Target	Status
Sarah Martinez	Lead K8s migration to production	100%	Q3 2024	Complete
Sarah Martinez	Mentor 2 junior engineers	100%	Q3 2024	Complete
James Liu	Design distributed cache system	75%	Q4 2024	In Progress
Alex Thompson	Complete 3 feature PRs independently	33%	Q3 2024	Behind

👥 What is HRFLOW?

HRFLOW (HR Performance Review Intelligence) is a multi-agent system that analyzes employee performance reviews and HR documents. It automatically parses review content, extracts development goals and competency ratings, analyzes feedback patterns, prioritizes HR actions, assigns follow-ups, and escalates concerns like PIP recommendations or retention risks — turning lengthy performance reviews into structured, actionable HR intelligence.

⚡ How It Works

Paste a performance review document. HRFLOW processes it through 6 specialized agents:

Step 1 — Extraction

📄 ReviewParser

Parses the performance review document — identifies employee info, ratings, competency scores, manager comments, self-assessment sections, and review period details.

Step 2 — Extraction

🎯 GoalExtractor

Extracts development goals, OKRs, and growth objectives — identifies skill gaps, training needs, career aspirations, and measurable targets set for the next review period.

Step 3 — Extraction

💬 FeedbackAnalyzer

Analyzes feedback sentiment and patterns — detects positive recognition, constructive criticism, recurring themes, and potential bias indicators across review narratives.

Step 4 — Decision

📈 TaskPrioritizer

Prioritizes HR actions by urgency — PIP recommendations and retention risks get Critical, while routine goal updates and training enrollments get lower priority.

Step 5 — Decision

👤 OwnerAssigner

Routes actions to the right stakeholder — development plans to managers, compensation reviews to HR, training requests to L&D, and policy concerns to HR leadership.

Step 6 — Decision

🚨 EscalationDecider

Escalates critical HR situations — performance improvement plans, retention risks for key talent, harassment or bias indicators, and compliance-sensitive review patterns.

🎬 Live Agent Pipeline Demo

Watch how HRFLOW processes a performance review through all 6 agents in real-time

Employee: Sarah Martinez Role: Senior Software Engineer
Rating: Exceeds Expectations Manager: David Kim
Strengths: Led migration of auth service with zero downtime. Mentored 2 juniors.
Growth Areas: Needs to improve cross-team communication. Missed Q2 OKR deadline due to scope creep.

📄

Step 1

ReviewParser

Idle

Rating: 4/5 · 6 competencies scored

🎯

Step 2

GoalExtractor

Idle

3 goals: communication, OKR, mentoring

💬

Step 3

FeedbackAnalyzer

Idle

Positive: 65% · Growth: 35%

📈

Step 4

TaskPrioritizer

Idle

P1: Communication · P2: OKR process

👤

Step 5

OwnerAssigner

Idle

Manager → Dev plan · L&D → Training

🚨

Step 6

EscalationDecider

Idle

Promotion candidate → HR review

✅ Pipeline Complete — 6 agents processed in 3.4s

3

Goals

65%

Positive

2

Assigned

★

Promo Ready

🎯 Real-World Use Cases

Annual Review Analysis

Process hundreds of performance reviews during annual review cycles — identify organization-wide trends, calibrate ratings consistency, and surface top performers.

Promotion Readiness Assessment

Evaluate promotion candidates objectively — track goal completion rates, competency growth trajectories, and multi-quarter performance trends.

PIP & Retention Risk Detection

Early detection of underperformance patterns and flight-risk indicators — enabling proactive interventions before losing key talent or legal escalations.

Development Goal Tracking

Track skill development and training completion across the organization — identify skill gaps, measure L&D program effectiveness, and plan workforce development.

🏗 Architecture

🔄

Orchestration Engine

Coordinates all 6 agents sequentially, processing reviews from parsing through escalation decisions.

🛡

Circuit Breaker Recovery

Per-agent fault isolation ensures goal extraction failures don't block feedback analysis or escalation decisions.

📋

Full Audit Trail

Every extracted goal, feedback pattern, and escalation is logged — essential for HR compliance and legal defensibility.

👥 Submit Performance Review LIVE AGENTS

📄

Drag & drop a review file, or click to browse

.txt .md supported

Employee Role

Live Agent Results

Extracted Goals & Development Items

Item	Category	Priority

— or analyze a sample review —

📄

Review
Parser

Extraction

Idle

›

🎯

Goal
Extractor

Extraction

Idle

›

💬

Feedback
Analyzer

Extraction

Idle

›

📈

Task
Prioritizer

Decision

Idle

›

👤

Owner
Assigner

Decision

Idle

›

🚨

Escalation
Decider

Decision

Idle

Analysis Log

Submit a performance review to begin analysis…

Goals Extracted

—

Feedback Items

—

Analysis Time

—

Total Tests

12

Passed

12

Failed

0

Duration

1.89s

Agent Tests

TestReviewParser

test_review_parsing0.20sPASS

test_empty_review0.07sPASS

TestGoalExtractor

test_goal_extraction0.17sPASS

TestFeedbackAnalyzer

test_feedback0.15sPASS

TestTaskPrioritizer

test_prioritization0.11sPASS

TestOwnerAssigner

test_assignment0.12sPASS

Infrastructure Tests

TestRecoveryStrategies

test_retry_strategy0.04sPASS

test_recovery_manager0.06sPASS

TestCircuitBreaker

test_circuit_closed0.03sPASS

test_circuit_opens0.04sPASS

TestAuditLogger

test_log_event0.07sPASS

test_audit_export0.10sPASS

Total Events

32

Reviews Processed

24

Avg Confidence

80%

Escalations

2

Review Audit Timeline

11:30:01 · workflow

Performance Review Analysis Started

Sarah Martinez — Senior Software Engineer

11:30:02 · agent

ReviewParser → Exceeds Expectations

4.5/5 overall rating · 3 peer reviews consolidated

11:30:03 · agent

GoalExtractor → 4 goals tracked

3/4 complete, 1 in progress · K8s migration, mentoring

11:30:04 · agent

FeedbackAnalyzer → Strongly Positive

Peer feedback highlights: leadership, technical depth, mentorship

11:30:05 · decision

Promotion Recommended: Senior → Staff Engineer

Based on sustained exceeds rating, completed goals, peer feedback

11:30:06 · workflow

Analysis Complete

4 goals, 5 feedback themes, promotion recommendation generated

Review Insights

⭐ Sarah Martinez — Exceeds Expectations

• Led K8s migration (on time, 40% cost reduction)
• Mentored 2 junior engineers (both now mid-level)
• Published internal tech talks (3 sessions, 200+ attendees)
• Zero production incidents in managed services

💡 Development Areas Identified

• Cross-team communication could improve
• Should delegate more rather than taking on everything
• Public speaking outside company would raise profile

🏆 Recommendations

• Promote to Staff Engineer (strong candidate)
• Assign as tech lead for next platform initiative
• Enroll in leadership development program
• Conference speaking opportunity (KubeCon 2025)

Circuit Breakers

6

All closed

Recoveries

1

Retry Rate

100%

Uptime

99.9%

Circuit Breaker Status

ReviewParser

CLOSED

0 failures

GoalExtractor

CLOSED

0 failures

FeedbackAnalyzer

CLOSED

0 failures

TaskPrioritizer

CLOSED

0 failures

OwnerAssigner

CLOSED

0 failures

EscalationDecider

CLOSED

0 failures