Forem: Hala Kabir

SkillMatch: Breaking the Glass Ceiling with Skill-First Logic By Hala Kabir

Hala Kabir — Sun, 05 Apr 2026 16:33:53 +0000

This is a submission for the 2026 WeCoded Challenge: Echoes of Experience.

🥋 The Journey: From the Dojang to the Dashboard

My name is Hala. I am 11 years old, a certified Python instructor, and a 1st Dan Taekwondo Black Belt. In martial arts, we are taught that our strength comes from our technique and discipline, not from who we are or where we come from.

When I started coding, I noticed that the tech industry doesn't always work that way. Sometimes, "unconscious bias" acts like a glass ceiling—a barrier that stops talented women and underrepresented voices before they even get a chance to speak. For the WeCoded 2026 Challenge, I decided to build a tool that turns that "Glass Ceiling" into a "Glass Dashboard"—transparent, beautiful, and focused entirely on talent.

🚀 The Project: SkillMatch

SkillMatch is a full-stack marketplace designed for Blind-Hiring. Instead of showing names, photos, or genders, the search results focus 100% on Tech Stacks.

How it works:

The Developer: Logs in and posts their specific skills (e.g., "FastAPI, React, Python").
The Recruiter: Searches the global board by skill. They see the "Talent Cards" first.
The Connection: Only when the recruiter clicks "Connect" do they see the person's name and open a direct WhatsApp link to chat.

🛠️ The Tech Stack

I wanted to build something professional, so I moved away from simple scripts and built a real Full-Stack API bridge.

Backend: Python 3.14 with the FastAPI library.
Frontend: HTML5 & JavaScript (Fetch API).
Design: Tailwind CSS with a "Glassmorphism" aesthetic.
Communication: WhatsApp Deep Linking API.

💻 Code Highlights

1. The "Equity" Search Logic (Python/FastAPI)

This backend function ensures that recruiters can only find people based on the skills they actually possess. It filters the data instantly without revealing the user's identity until requested.

@app.get("/search")
def search_talent(skills: str = ""):
    # If no search query, show everyone in the talent pool
    if not skills:
        return talent_pool

    # Filter the list based on skill keywords (Case-insensitive)
    results = [p for p in talent_pool if skills.lower() in p["skills"].lower()]
    return results

2. The "Bridge" (JavaScript & WhatsApp API)

I used Deep Linking to make the connection between the recruiter and the developer feel instant and personal.

// This creates a personalized WhatsApp link when a recruiter finds a match

async function runSearch() {
    const res = await fetch(`${API}/search?skills=${query}`);
    const data = await res.json();

    grid.innerHTML = data.map(p => {
        // We pre-fill the message so the developer knows exactly why they are being contacted
        const message = encodeURIComponent(`Hi ${p.name}, I saw your ${p.skills} skills on SkillMatch! Ready to build?`);
        const waLink = `https://wa.me/${p.whatsapp}?text=${message}`;

        return `
            <div class="card">
                <h4>${p.skills}</h4>
                <a href="${waLink}" class="connect-btn">Connect with ${p.name}</a>
            </div>
        `;
    }).join('');
}

💡 Lessons Learned

Building this at 11 years old taught me that code is a superpower. You can use it to build walls, or you can use it to build doors. By using FastAPI and Tailwind, I learned how to handle real data and create a UI that feels modern and welcoming.

The biggest lesson? Equity isn't just a goal; it's a design choice. When we choose to build apps that prioritize skills over names, we choose to make the tech industry a fairer place for everyone.

🏁 Conclusion

I hope SkillMatch inspires other young developers to build tools that solve real problems. Whether you are a Black Belt in a martial art or a Black Belt in Python, your contribution to tech matters.

Let's break some barriers together! 🥋💻

Auto-Researcher Hub: Bridging AI Knowledge and Workspace Automation By Hala Kabir

Hala Kabir — Sat, 28 Mar 2026 10:57:27 +0000

This is a submission for the Notion MCP Challenge

What I Built:

The Auto-Researcher Hub is an agentic workspace architect designed to solve the "blank page" problem in high-velocity research. As an AI and Blockchain developer, I often find myself overwhelmed with technical papers and scattered documentation.

I built a system that doesn't just "talk" about research but physically builds the environment for it. Using a Python-based orchestration layer, the tool automatically constructs a structured Research Database in Notion, complete with custom properties (Summary, URL, Status), and populates it with verified data points on Multi-Agent Systems and ZK-Proofs.

Video Demo:

Watch the Auto-Researcher Hub in action here!

Watch the Auto-Researcher Hub in action here!
youtube : click here

X: click here

Linkdin: click here

The full source code:

The full source code, including the orchestration script and the README, can be found here:

GitHub Repository: notion-auto-researcher

How I Used Notion MCP:

Integrating the Notion Model Context Protocol (MCP) was a game-changer for my workflow. Traditionally, chatbots are "trapped" in a chat bubble; they can tell you what to do, but they can't do it for you.

By using the Notion MCP server (specifically @modelcontextprotocol/server-notion), I unlocked the ability for an AI agent to:

Architect Schemas:

Use Python logic to define exactly what a "Research Database" should look like.

Contextual Awareness:

The agent can "see" my existing Notion pages and decide where to build the hub.

Automated Population:

It bridges the gap between raw web data and structured productivity, turning a manual 15-minute setup task into a 5-second automated execution.

This integration transforms Notion from a passive notebook into an active, collaborative partner that can organize my thoughts before I even start typing.

--By Hala Kabir

Learning MongoDB isn’t just about storing data. It’s about understanding how modern applications think about information.

Hala Kabir — Fri, 06 Mar 2026 02:31:11 +0000

Learning MongoDB isn’t just about storing data.

Hala Kabir ・ Mar 6

#programming #mongodb #vectordatabase #webdev

Learning MongoDB isn’t just about storing data.

Hala Kabir — Fri, 06 Mar 2026 02:30:07 +0000

[Boost]

Hala Kabir ・ Mar 6

#programming #mongodb #vectordatabase #webdev

[Boost]

Hala Kabir — Fri, 06 Mar 2026 02:28:48 +0000

Getting Started with MongoDB: The Database Built for Modern Applications 🍃🚀 by Hala Kabir

Hala Kabir ・ Mar 6

#programming #mongodb #vectordatabase #webdev

Getting Started with MongoDB: The Database Built for Modern Applications 🍃🚀 by Hala Kabir

Hala Kabir — Fri, 06 Mar 2026 02:17:34 +0000

Ever wondered how modern apps store massive amounts of flexible data without breaking the system?

Traditional databases store data in strict tables.

But modern applications need something faster, more flexible, and scalable.

That’s where MongoDB comes in.

Instead of rows and columns, MongoDB stores data in documents that look like JSON.
This makes it incredibly popular for web apps, AI systems, and scalable platforms.

Let’s explore how it works 👇

🌿 What Makes MongoDB Different?

Traditional SQL databases look like this:

id	name	age
1	Hala	18

But MongoDB stores data like this:

`{
  "name": "Hala",
  "age": 18,
  "skills": ["Python", "AI", "Taekwondo"]
}
`

See the difference?

MongoDB allows dynamic and flexible data structures, which is perfect for modern development.

🧠 Think of MongoDB Like a Smart Digital Library

Imagine a library.

Instead of forcing every book to follow the same format, the library allows each book to contain different chapters, notes, and sections.

That’s exactly how MongoDB treats data.

It stores information as documents inside collections, rather than rows inside tables.

Structure:

Database
   ↓
Collection
   ↓
Documents

Example:

School Database
     ↓
Students Collection
     ↓
Student Documents

⚡ Step 1: Creating Your First Document

Let’s insert a simple user document.

db.users.insertOne({
  name: "Hala",
  role: "AI Developer",
  country: "Pakistan"
})

Boom 💥
You just stored data in MongoDB.

🔍 Step 2: Reading Data

To retrieve data:

db.users.find()

Want a specific user?

db.users.find({name: "Hala"})

MongoDB queries are simple and powerful.

✏️ Step 3: Updating Data

Let’s update a user role.

db.users.updateOne(
  { name: "Hala" },
  { $set: { role: "AI Engineer" } }
)

MongoDB updates only the fields you specify.
Efficient and clean.

❌ Step 4: Deleting Data
Need to remove a record?

db.users.deleteOne({name: "Hala"})

Just like that — the document disappears.

🚀 Why Developers Love MongoDB

Here’s why MongoDB became one of the most popular databases in modern tech:

✅ Flexible schema
✅ Perfect for rapid development
✅ Scales easily for large applications
✅ Works beautifully with JavaScript and APIs
✅ Ideal for AI, cloud apps, and microservices

Companies like Netflix, Uber, and eBay have used MongoDB in different parts of their infrastructure.

🧪 Mini Project Idea (For Beginners)

Try building a Student Database System.

Your collection might look like this:

{
  "name": "Ali",
  "grade": 10,
  "subjects": ["Math", "Physics", "Computer Science"]
}

Then practice:

Insert students

Update grades

Search by subject

Delete records

This simple exercise teaches you real database thinking.

🧠 What Learning MongoDB Actually Teaches You

Working with MongoDB builds key backend skills:

🔹 Data modeling
🔹 Query optimization
🔹 CRUD operations
🔹 API integration
🔹 Scalable system design

These concepts power:

Social media apps

AI systems

Cloud platforms

Enterprise software

🔥 Want to Level Up?

Once you're comfortable, try exploring:

⚡ MongoDB Atlas for cloud databases
⚡ Building APIs with Node.js
⚡ Using MongoDB with Python and FastAPI

This combination powers real production systems.

🎯 Final Thought

Learning MongoDB isn’t just about storing data.

It’s about understanding how modern applications think about information.

Small experiments today can lead to powerful systems tomorrow.

If you can design your data well,
you can build almost any application.

Happy building 🍃💻

How I Built ArtAura: An AI Art "Sensei" to Help Artists Beat the Blank Page

Hala Kabir — Sat, 28 Feb 2026 19:37:09 +0000

This is a submission for the DEV Weekend Challenge: Community

The Community

I built this for the Global Artist Community. As an 11-year-old who loves both drawing and coding, I realized that we all face the same monster: The Blank Page. Sometimes I sit down to create, but I just don't know what to draw. I wanted to build a home for artists where they never have to guess. It’s a space for people who want a weekly reason to pick up their pencil and a fun way to get feedback on their progress.

What I Built

ArtAura is a 20-week interactive art journey. It’s a "Studio" app that:

Challenges You: Gives a new, creative prompt every single week.
Evaluates with AI: Uses a "Sensei" (Gemini AI) to look at your art and give you a score out of 10 plus a tip to improve.
Celebrates Wins: When you hit a perfect 10/10, the app throws a party! You get a special cupcake, a golden "Congratulations" banner drops down, and colorful confetti rains over your screen.
Interactive Fun: I even added a feature where you have to press the "F" key to "blow out" the candle on your reward cupcake!

Demo

You can see ArtAura in action here:

Code

I used Vanilla JavaScript, HTML5, and CSS3. I integrated the Uploadcare API for the image uploads and the Gemini API for the AI grading.

How I Built It

Building this was a fun puzzle!

Glassmorphism Design

I wanted the app to feel modern, so I used semi-transparent "glass" cards with blur effects.

The "Z-Index" Battle

One of the hardest parts was making sure the confetti appeared in front of the celebration screen. I had to learn how to manipulate layers in JavaScript so the "Confetti Rain" wouldn't hide behind the cupcake!

The Reward System

I used the canvas-confetti library for the rain effect and CSS transitions for the "Golden Banner" that drops from the top of the screen.

I’m really proud that I could combine my two favorite hobbies—drawing and programming—to help other artists get over their creative blocks!

Multi-Agent-System-Gemini by Hala Kabir

Hala Kabir — Wed, 25 Feb 2026 21:13:26 +0000

[Boost]

Hala Kabir ・ Feb 25

#agents #buildmultiagents #gemini #adk

[Boost]

Hala Kabir — Wed, 25 Feb 2026 21:12:44 +0000

🚀 Building a Multi-Agent Content Studio with Gemini 2.5 This post is my submission for .

Hala Kabir ・ Feb 25

#agents #buildmultiagents #gemini #adk

🚀 Building a Multi-Agent Content Studio with Gemini 2.5 This post is my submission for .

Hala Kabir — Wed, 25 Feb 2026 20:39:07 +0000

*This post is my submission for [DEV Education Track: Build

GitHub Multi-Agent-System-Gemini
🌟 What I Built

I developed a Multi-Agent AI Content Studio designed to solve the biggest problem every writer faces:
Writer's Block. Instead of just asking a single AI to "write a story," I built a collaborative system where three specialized AI "workers" (Agents) talk to each other. Each agent has a specific job, a specific personality, and a specific goal. By breaking the work into pieces, the final article is much more accurate, better researched, and sounds more human.

🧠 Why Multi-Agent Systems?

As an 11-year-old learning to code, I realized that even the smartest AI can sometimes get overwhelmed if you ask it to do too much at once.

Think of it like a professional kitchen: You don't have one person doing everything. You have a Chef to plan, a Sous-Chef to chop, and a Server to check the plate. My system works exactly like that. It ensures that the facts are checked before the writing starts, and the writing is polished before the user sees it.

**
🤖 Meet the Agents**

I used the latest Google GenAI SDK and the Gemini 2.5 Flash model to power my team. Here is the breakdown of my squad:

1. The Research Agent (The Brains)

This agent’s job is to gather the "building blocks." It ignores the "fluff" and focuses on key points, subtopics, and relevant examples. It makes sure the content is grounded in good information.

2. The Writer Agent (The Creator)

This agent receives the research and looks at the requested Tone (Professional, Casual, or Funny). It’s responsible for the structure, the headings, and making the information interesting to read.

3. The Reviewer Agent (The Perfectionist)

This is the final gatekeeper. It reads the draft from the Writer Agent and looks for ways to improve the flow, simplify hard sentences, and make sure the article is engaging from start to finish.

💻 The Logic Behind the System

I used Python and Streamlit to build a clean user interface. The most important part of the code is the Orchestrator. This is the master function that coordinates the "hand-offs" between agents.

🧗 The Challenges (My Debugging Journey)

Building this wasn't easy! I hit several walls that almost made me stop, but I pushed through:

The 404 Model Mystery:

At first, I kept getting "Model Not Found" errors. I learned that AI versions change fast! I had to update my code from Gemini 1.5 to the brand new Gemini 2.5 to get it to work.

The Permission Puzzle:

I had to learn how to manage .env files and API keys securely. I realized that keeping your keys secret is the first rule of being a real developer.

Agent Hand-offs:

Making sure the Reviewer Agent actually received the text from the Writer Agent was a lesson in how functions pass data in Python.

📚 Key Learnings

This project taught me that being a developer isn't just about writing code; it's about problem-solving. I learned that:

Modularity is key:

Breaking a big problem into small agents makes the code easier to fix.

Prompt Engineering matters:

How I talk to the agents changes how they perform. I had to give them very clear "instructions" to make them behave.

*The AI Community is great: *

Using documentation and learning from tracks like this one on DEV.to helps you grow faster.

Thanks for checking out my project! I can't wait to see what other multi-agent systems people are building. 🚀

By Hala Kabir The youngest certified AI & Blockchain Professional App & Chatbot Developer

What if two AI models competed against each other… and became better by trying to defeat one another? That’s exactly how GANs work. And they changed AI forever.

Hala Kabir — Sun, 15 Feb 2026 18:55:46 +0000

GANs Explained Simply: The Two-Neural-Network Battle That Changed AI

Hala Kabir ・ Feb 15

#ai #machinelearning #python #deeplearning

[Boost]

Hala Kabir — Sun, 15 Feb 2026 18:51:13 +0000

GANs Explained Simply: The Two-Neural-Network Battle That Changed AI

Hala Kabir ・ Feb 15

#ai #machinelearning #python #deeplearning