Performance Review Agent

Agent Description:

The Performance Review Agent is a multi-stage AI orchestration system designed to generate objective, evidence-based employee evaluation reports. By integrating raw performance data with internal HR frameworks and industry benchmarks, it ensures that every review is fair, policy-compliant, and data-driven.

This agent enables organizations to:

  • Eliminate Evaluation Bias: Rely on structured KPI achievement data rather than subjective manager recall.

  • Automate Policy Application: Instantly determine promotion eligibility or PIP requirements by cross-referencing internal HR PDF manuals.

  • Synthesize Qualitative Feedback: Summarize peer and manager comments into actionable behavioral insights.

  • Benchmark against Industry: Compare employee performance against broader industry standards to identify top talent or "at-risk" employees.

  • Scale HR Operations: Generate professional, audit-ready review drafts in seconds, allowing managers to focus on development coaching.

Purpose and Components
  • Purpose: Manual performance evaluations are slow, inconsistent, and biased, making it hard for HR to combine KPIs and feedback while ensuring policy‑aligned reviews.

    This agent automates the evaluation process by aggregating KPI metrics, standardizing qualitative feedback, and applying policy-aligned scoring rules. It ensures consistent, unbiased assessments, flags high performers and at‑risk employees, and generates policy‑compliant recommendations for promotions or PIPs. This creates faster, fairer, and more accurate performance evaluations.

  • Components:

    • Performance Data Extraction Agent: Ingests raw metrics and observations from PDFs to create structured datasets.

    • Policy & Benchmark Evaluation Agent: Applies internal frameworks and industry standards to determine objective ratings.

    • Trend and Risk Analysis Agent: Evaluates performance trajectories over multiple quarters to identify high-potential or at-risk employees.

    • Performance Review Drafting Agent: Synthesizes all findings into a structured, unbiased final report.

Supported Capabilities

  • Eliminate Evaluation Bias: Relying on structured KPI achievement data instead of subjective recall.

  • Automate Policy Application: Instantly determining promotion or PIP eligibility by cross-referencing internal HR manuals.

  • Qualitative Feedback Synthesis: Summarizing peer and manager comments into actionable behavioral insights.

  • Industry Benchmarking: Comparing individual performance against broader industry standards.

  • Scalable HR Operations: Generating professional, audit-ready review drafts rapidly to allow managers to focus on coaching.

LLM Used

  • OPENAI GPT_4O_MINI

    Note: To learn more about the LLM and to modify its behavior, refer to the Configuring LLM settings section.

Sub-Agents

1. Performance Data Extraction Agent

  • Role:Data Extractor

  • Scope:Extracts metrics, feedback, and observations from uploaded PDFs without interpretation.

  • Description:Ingests raw KPI reports, peer reviews, and manager notes. It parses these documents to create a structured dataset of quantitative targets vs. achievements and qualitative strengths/weaknesses.

  • LLM Used: Default (Inherits from parent agent).

2. Policy and Benchmark Evaluation Agent

  • Role:Policy Evaluator

  • Scope:Applies the internal evaluation framework, promotion rules, and industry benchmarks.

  • Description: A REACT agent that utilizes an MCP Client to read internal HR policy documents. It calculates weighted performance scores, determines official ratings (1–5), and checks for promotion or PIP eligibility based on hard-coded policy logic.

  • LLM Used: Default (Inherits from parent agent).

3. Trend and Risk Analysis Agent

  • Role:Risk Analyzer

  • Scope:Detects multi-quarter trends and triggers logic for PIP or succession planning.

  • Description:Analyzes whether an employee's performance is on an upward or downward trajectory over multiple quarters. It identifies recurring behavioral risks and highlights high-potential indicators to recommend either a PIP or a promotion track.

  • LLM Used: Default (Inherits from parent agent).

4. Performance Review Drafting Agent

  • Role:Report Generator

  • Scope:Generates the final, structured, and unbiased performance review draft.

  • Description:Synthesizes data from the previous three agents to draft a comprehensive report. It includes an Executive Summary, KPI overview, Benchmark comparison, and specific action plans for development, ensuring a fully objective tone.

  • LLM Used: Default (Inherits from parent agent).

Tools Used:

  • MCP Client (RAG): Queries internal HR PDFs including "Promotion Policy" and "Evaluation Frameworks" to ensure compliance.

  • Google Search: Captures real-time industry benchmark standards for specific job roles to provide external context.
Note: For details on modifying the Tools, refer Tools Library section.
Agent Workflow Behavior Summary

  1. Input: Manager uploads an employee’s quarterly KPI sheet and peer feedback forms.

  2. Data Extraction:

    • The Extraction Agent pulls: "Sales Target: 105% achieved," "Peer Feedback: Strong collaborator," and "Attendance: 100%."

  3. Policy & Benchmark Audit:

    • The Evaluation Agent checks the Promotion_Policy.pdf.

    • Condition: If Achievement > 100% for 2 quarters -> Flag as Promotion Eligible.

    • It also checks the Benchmark_Guide.pdf to see how a 105% sales achievement compares to the market average.

  4. Trend Assessment:

    • The Risk Agent observes that performance has improved by 15% since Q1, marking the trajectory as "High Growth."

  5. Final Drafting:

    • The Drafting Agent assembles these facts into a professional report with a final rating and a justification for promotion.

Sample Questions:
  • Based on the uploaded performance report and company evaluation policies, generate a complete performance review including rating, benchmark comparison, and promotion or PIP recommendations <upload pdf>