From Skeleton to Production: Building HR Goal Tracking Portal with GitHub Copilot

This is a submission for the GitHub Finish-Up-A-Thon Challenge

What I Built

I built the Atomberg Goal Setting & Tracking Portal — a full-featured,
role-based HR performance management system for Atomberg Technologies (a
fast-growing Indian consumer electronics brand).

The portal manages the complete employee performance lifecycle:

• 🎯 Goal Setting — Employees create goals with weightage, UoM type, and thrust area alignment. Business rules enforced: max 8 goals, total 100% weightage, minimum 10% per goal
• ✅ Manager Approval Workflow — Submit → Review → Approve / Return for Rework
• 📊 Quarterly Check-ins — Q1–Q4 progress logging with 4 UoM types (Numeric Min/Max, Timeline, Zero-is-best)
• 🚨 Escalation Monitor — Rule-based detection of overdue submissions, missing approvals, and pending manager reviews
• 📈 Analytics & Reports — QoQ trend charts, department-wise performance, thrust area distribution, CSV exports
• 🔒 Admin Governance — Cycle phase management, shared goal distribution, full audit logging

Tech Stack: React 18 + Vite · Context API · localStorage (zero-backend
demo) · Recharts · Lucide Icons · Custom dark glassmorphism CSS

Demo

🔗 Live App: https://atomberg-goal-tracker.vercel.app/

📁 GitHub Repo: https://github.com/anonomous29/atomberg-goal-tracker

Quick Login Credentials

Role	Email	Password
Admin	[e-mail]	admin123
Manager	[e-mail]	manager123
Employee	[e-mail]	emp123

What to try:

1. Login as Admin → Go to Escalation Monitor → see rule-based alerts
2. Go to Reports → QoQ Trends tab → see Q1 performance bar chart
3. Login as Employee → Dashboard shows Q1 score donut + goal-level progress
4. Login as Manager → See team check-in status and review Q1 feedback

The Comeback Story

This project started as a hackathon requirement — a Business Requirements
Document (BRD) from Atomberg asking for a digital goal management system. The
initial version had the core structure but was essentially an empty shell:
blank dashboards on login, no check-in data, no analytics, and missing two
bonus modules entirely.

What I finished up:

Before

• ❌ Empty dashboards on every login (no seed data)
• ❌ No Escalation Monitor (BRD §5.3 bonus feature)
• ❌ Basic analytics — just one bar chart
• ❌ No QoQ trend analysis
• ❌ Manager dashboard didn't show check-in status
• ❌ No per-goal score breakdown in reports

After

• ✅ Rich pre-seeded demo data — 16 approved goals, 4 Q1 check-ins with manager comments, 13 audit entries, 8 notifications loaded on first open
• ✅ Escalation Monitor — 5 rule types (Critical/High/Medium/Low), collapsible employee cards, department filter, CSV export
• ✅ QoQ Analytics — 4-tab reports page with goal-level achievement table, quarter-on-quarter grouped bar chart, department comparison, and pie distribution
• ✅ Score donut charts on employee dashboard
• ✅ Team check-in tracking on manager dashboard (submitted vs. reviewed)
• ✅ Version-aware localStorage — auto-migrates returning users to new seed data

The biggest unlock was the Escalation Monitor — it transformed the portal
from a passive data entry tool into an active compliance tracker that tells
managers exactly who is falling behind and why.

My Experience with GitHub Copilot

GitHub Copilot was integral to finishing this project quickly. Here's
specifically how it helped:
1. Boilerplate elimination
The score computation engine, localStorage CRUD helpers, and recharts
configurations were tedious to write from scratch. Copilot autocompleted
entire function bodies after seeing the first few lines of pattern.
2. Business logic translation
The BRD had complex UoM scoring rules (Numeric Min, Numeric Max, Timeline,
Zero). Copilot helped translate the prose spec directly into the computeScore
switch statement — getting the inverted formula for "lower is better" metrics
correct on the first try.
3. React pattern consistency
Across 10+ components, Copilot kept prop patterns, className conventions, and
inline style objects consistent — no more copy-paste drift between components.
4. Escalation rule engine
The most valuable assist: when I described the escalation rules in a comment
block, Copilot generated the full predicate functions for each rule type,
including edge cases like "employee has goals but hasn't submitted them."
5. Seed data generation
Writing 16 realistic performance goals with correct weightages summing to 100%,
meaningful descriptions, and plausible Q1 achievement numbers would have taken
an hour manually. Copilot drafted the full INITIAL_GOALS and
INITIAL_CHECKINS arrays in minutes.
Overall, Copilot cut the "finishing up" time from an estimated 2 days to a
single focused session — letting me ship the escalation module, analytics
overhaul, and rich demo data all in one go.

Atomberg Goal Tracker — Demo Screenshots

Screen 1: Admin Dashboard

Admin Dashboard
Organization-wide overview — team stats, department progress, quick actions

Screen 2: Escalation Monitor (Bonus Feature)

Escalation Monitor
Rule-based compliance tracker — Critical / High / Medium / Low severity alerts

Screen 3: QoQ Analytics — Reports & Analytics

QoQ Trends
Quarter-on-Quarter performance trends — all employees compared across Q1–Q4

Screen 4: Employee Dashboard

Employee Dashboard
Personal dashboard with Q1 score donut, approved goals, and cycle timeline

1. Boilerplate elimination
The score computation engine, localStorage CRUD helpers, and recharts
configurations were tedious to write from scratch. Copilot autocompleted
entire function bodies after seeing the first few lines of pattern.

2. Business logic translation
The BRD had complex UoM scoring rules (Numeric Min, Numeric Max, Timeline,
Zero). Copilot helped translate the prose spec directly into the computeScore
switch statement — getting the inverted formula for "lower is better" metrics
correct on the first try.

3. React pattern consistency
Across 10+ components, Copilot kept prop patterns, className conventions, and
inline style objects consistent — no more copy-paste drift between components.

4. Escalation rule engine
The most valuable assist: when I described the escalation rules in a comment
block, Copilot generated the full predicate functions for each rule type,
including edge cases like "employee has goals but hasn't submitted them."

5. Seed data generation
Writing 16 realistic performance goals with correct weightages summing to 100%,
meaningful descriptions, and plausible Q1 achievement numbers would have taken
an hour manually. Copilot drafted the full INITIAL_GOALS and
INITIAL_CHECKINS arrays in minutes.

Overall, Copilot cut the "finishing up" time from an estimated 2 days to a
single focused session — letting me ship the escalation module, analytics
overhaul, and rich demo data all in one go.g! -->