SafeHire AI

AI-Powered Career Re-entry Assistant for Women
Built for SheBuilds Chennai Hack — Code & Challenge 3.0
🎯 SDG 5: Gender Equality | Rajalakshmi Engineering College, Chennai

---

🌟 What is SafeHire AI?

SafeHire AI is a web platform that helps women confidently return to the workforce after career breaks — whether due to maternity, caregiving, health, or personal reasons. It provides personalized AI-powered guidance, skill roadmaps, resume support, and returnship-ready job opportunities — all in one place. <<<<<<< HEAD

---

✨ Features

Feature	Description
📄 Resume Gap Analyzer	Identifies skill gaps and rewrites career break narrative positively
🗺️ Personalized Career Roadmap	Step-by-step milestones based on break duration and target role
💼 Returnship Finder	Curated companies with active returnship programs
🎤 AI Interview Coach	Practice rounds with feedback and confidence scoring
📊 Confidence Booster Dashboard	Visual progress tracking with animated score rings
🤝 Community & Mentorship	Connects women with mentors, coaches, and peer returners
🧠 Emotional Readiness Check ⭐	Detects imposter syndrome patterns and serves targeted support
⏱️ Break Duration Intelligence ⭐	Dynamically adjusts all advice based on exact break length
🛠️ AI Skill Sandbox Simulator ⭐	Modernizes older skills to match 2026 industry standards

⭐ = Unique to SafeHire AI — not found in existing platforms

---

🎨 UI Highlights

• Blue + pink gradient theme with elegant typography
• Dark / Light mode toggle (persists across pages)
• Animated circular score rings (Skill Readiness, Resume Strength, Interview Confidence)
• Skill breakdown bars — current skills vs skills to learn
• Flip cards on landing page for feature previews
• 4-step multi-step onboarding form with resume upload
• Fully responsive design

---

🛠️ Tech Stack

• Frontend — Pure HTML, CSS, JavaScript (no framework)
• AI — Gemini API (gemini-2.5-flash), called directly from the browser
• Fonts — Playfair Display, DM Sans, Space Grotesk (Google Fonts)
• Deployment — GitHub Pages / Live Server

---

🚀 Setup & Run

1. Clone the repo

git clone https://github.com/YOUR_USERNAME/SafeHire-AI.git
cd SafeHire-AI

2. Add your Gemini API key

Get a free API key from Google AI Studio.

In dashboard.html, find the line near the top of the <script> section:

const GEMINI_API_KEY = "YOUR_GEMINI_API_KEY";

and replace it with your own key.

⚠️ Note: this is a client-side key for hackathon/demo purposes only. Don't reuse this key elsewhere, since it's visible to anyone viewing the page source.

3. How the AI calls work

All AI logic lives inline inside dashboard.html. Each of the 9 feature cards calls Gemini directly from the browser:

const GEMINI_API_KEY = "YOUR_GEMINI_API_KEY";
const GEMINI_URL = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent?key=${GEMINI_API_KEY}`;

async function askgemini(prompt, resultId) {
  const el = document.getElementById(resultId);
  try {
    const response = await fetch(GEMINI_URL, {
      method: "POST",
      headers: { "Content-Type": "application/json" },
      body: JSON.stringify({
        contents: [{ parts: [{ text: prompt }] }]
      })
    });

    const data = await response.json();

    if (!data.candidates) {
      el.innerHTML = `<div class="result-content" style="color:var(--pink)">⚠️ ${data.error?.message || 'No response from AI'}</div>`;
      return;
    }

    const answer = data.candidates[0].content.parts[0].text;
    el.innerHTML = `<div class="result-content">${answer.replace(/\n/g, '<br/>')}</div>`;

  } catch (err) {
    el.innerHTML = `<div class="result-content" style="color:var(--pink)">⚠️ ${err.message}</div>`;
  }
}

Each feature card builds a tailored prompt using the user's profile (name, target role, break duration, etc.) and calls askgemini(prompt, resultId) to populate its result section. Calls are spaced out slightly to stay within Gemini's free-tier rate limits.

4. Open the app

Open index.html with Live Server in VS Code

---

📁 Project Structure

SafeHire-AI/
├── index.html          # Landing page with flip feature cards
├── onboarding.html     # 4-step user profile form
├── dashboard.html      # AI results + visual score dashboard (Gemini calls live here)
├── style.css           # Full theme, dark mode, animations
├── script.js           # Reference/test file for AI logic experiments
└── README.md