Machine Learning · People Analytics · Product Build

HR Analytics Intelligence Platform

A HR analytics product built into a Python analytics application and mirrored inside this Cloudflare portfolio as a polished browser sandbox. The project brings together dashboarding, predictive HR analytics, exploratory analysis, offline Local AI, and optional advanced AI-assisted insight generation.

Application flow

The build follows the same logic as the Python analytics app: load data, filter automatically, generate visuals, explore relationships, run HR predictions, then ask AI-style questions against the active data.

01Load HR Data

Use sample files or upload CSV data.

02Filter Automatically

Checkbox selections update the dashboard instantly.

03Generate Visuals

Auto charts, KPIs and insights are built from the selected dataset.

04Predict & Explore

Flight risk, headcount forecast, pay equity and chart builder.

05Ask AI

Local AI answers with tables and visuals; advanced AI adds narrative depth.

Modules included

Each module is intentionally positioned as a portfolio-ready product area, not just a code demo.

Dashboard

📊 Executive HR View

Active dataset selection, checkbox filters, KPI cards, auto-generated insights, auto charts, summary statistics and raw data preview.

Prediction

🔮 HR Predictive Analytics

Flight Risk scoring, Headcount Forecast and Pay Equity review with clean KPIs, risk tiers, feature views and exportable tables.

Explore

🔧 Self-Service Analysis

Group-by, metric, aggregation, chart type, optional colour dimension and scatter Y-axis controls for flexible analysis.

Offline AI

💬 Local AI

Rule-based assistant that interprets HR questions, selects the right dataset, detects intent and returns text, table and visual outputs.

LLM

🤖 Advanced AI

Optional advanced AI assistant workflow for executive summaries, root-cause interpretation, recommendations and deeper people analytics narratives.

Deployment

☁️ Cloudflare-ready demo

Static browser sandbox for Cloudflare Pages, structured to demonstrate the analytics product without exposing backend package downloads on the public page.