Youtube:https://youtu.be/sjceSEYRj9o
| Photo | Name | Position | Social Media |
|---|---|---|---|
 |
Tuba İLHAN HORUZ | Scrum Master |
|
 |
Duygu Başak ACAR | Developer |
|
 |
Melih Taha BEKTAŞ | Developer |
|
 |
Saim Berk AKÇEŞME | Developer |
|
This project is a career discovery platform that allows high school seniors to consciously make their career choices with the support of big data, learning engineering and artificial intelligence.
What is Career Path?
Career Path is a platform that supports young individuals in making decisions that are sensitive not only to their own interests and abilities but also to societal needs, economic sustainability and environmental impacts when choosing their future careers.
New Generation Career Approach
Today's young people are choosing not just a profession, but also a lifestyle, a set of values, and a sphere of influence. Environmental crises, digital transformation, and structural changes in the workforce are among the most important factors shaping the professions of the future. At Career Path, we act with awareness of this reality.
This chart shows the team's progress and remaining workload during Sprint 1. Although progress was slow in the first days, the pace increased in the following days, and all work was successfully completed by the end of the sprint.
- Actual Remaining Work: Shows the amount of work remaining for the team at the end of each day.
- Ideal Progress: Shows what the reduction would be if an equal amount of work were completed each day.
-
Sprint Notes:
-It was decided to use
Mirofor project management. The Miro theme was created by Tuba İlhan Horuz.-It was decided to open a GitHup repo link. The Githup repo was created by Melih Aktaş.
-It was decided to use
Emailfor the login system.-It was decided to use
Wordpress (Custom HTML5 & Custom CSS3)for the web application.-For security, it was decided to use
Google reCAPTCHA + WP 2FA.-For social login, it was decided to use
Nextend Social Login.-For artificial intelligence integration, it was decided to use
Python API + REST API (secured with JWT).-For portable, independent infrastructure (especially when working with AI models), it was decided to use
Docker.-For the career recommendation algorithm (classification/recommendation engine), it was decided to use
scikit-learn & XGBoost.-For deep learning and data analysis and manipulation, it was decided to use
Pandas & NumPyandTensorFlow & PyTorch.-For powerful relational databases, it was decided to use
SQLandPostgreSQL.-The tests currently used for personality analysis were reviewed, and the final decision on which test to use was left to the second sprint phase.
-It was decided that occupational data would be collected primarily through the
O*NETwebsite. -
Estimated Score to Complete: 200 points.
-
Prediction Logic:
-
End of Sprint 1 Estimation Logic The tasks completed in Sprint 1 and the points assigned to them were determined in accordance with Google Project Management training, taking into account each task's complexity, effort, and importance to the project. To achieve the 200-point total, the following tasks and points were assigned:
Completed Tasks and Point Distribution:
Installing Miro and Creating a Task Board: 25 Points
Preliminary Project File Preparation and Updates: 35 Points
Creating a GitHub Repo: 20 Points
Deciding on the Application and Team Name: 20 Points
Creating a Logo and Slogan: 30 Points
WordPress Installation and Theme Selection: 40 Points
Creating a WhatsApp Group and Community, File Backup: 15 Points
Team Meetings and Active Communication: 15 Points
Total Points: 25 + 35 + 20 + 20 + 30 + 40 + 15 + 15 = 200 Points.
Target Point: 200 Points.
This score reflects the weight of the core organizational and initial technical tasks completed in Sprint 1. Target score for Sprint 1 has been achieved.
-
Daily Scrum:
-
Meetings were held daily via Google Meet.
-
Active communication was maintained via the WhatsApp group.
-
 |
 |
|---|---|
| WhatsApp Group - Screenshot 1 | WhatsApp Community- Screenshot 2 |
- Product Backlog URL: https://miro.com/app/board/uXjVIicQLWg=/?share_link_id=476687804339 (Miro)
- Sprint Review:
- Miro was installed for the project and the team task board was created.
- We prepared the project file and made updates to it.
- We had difficulty deciding on an application name. We were torn between 'PathPilotAI' and 'CAREERPATH', so we chose the team name 'PathPilotAI' and the project name 'CAREERPATH', taking another step towards branding.
- We hadn't decided on a color palette and hadn't yet finalized the logo.
- Tuba İlhan Horuz created the logo and slogan.
- Melih Talha Aktaş installed WordPress and selected the theme.
- It was decided that the data to be used about professions would be collected from the O*NET website.
- We researched existing personality tests.
Overall, we believe we had a good sprint process. The sprint process was as planned. Because our team was formed later, we held daily Google Meet meetings to complete all our planning within a week. We actively communicated through our WhatsApp group. We created a WhatsApp community for the group to back up documents and work files.
-
Sprint Review Participants:
Tuba İlhan Horuz,Duygu Başak Acar,Sude Nur Kömür,Melih Talha Bektaş,Saim Berk Akçeşme -
Sprint Retrospective:
-
At the team meeting in the second sprint, it was decided that only Melih and Sude Nur would write code for the website.
-
In the second sprint, we decided to prepare the Docker environment and perform container orchestration.
-
In the second sprint, we will install the PostgreSQL database, perform backups, and configure roles.
-
In the second sprint, we will implement custom JavaScript integrations for dynamic sections in WordPress.
-
In the second sprint, we decided to implement a Backend Python environment (FastAPI) and dependency management.
-
In the second sprint, we decided to establish a REST API bridge and a JWT-based secure connection.
-
In the second sprint, we decided to install the relevant libraries (Pandas & NumPy) for data analysis.
-
In the second sprint, we decided to install scikit-learn & XGBoost and develop a sample career recommendation model.
-
It was decided to use
TensorFlow & PyTorchto create a deep learning environment in the second sprint. -
In the second sprint, it was decided to implement the test to be used for personality analysis and integrate an NLP scorer.
-
In the second sprint, it was decided to install the WP User Manager Ultimate Member plugin.
-
In the second sprint, it was decided to add the Google reCAPTCHA + WP 2FA security plugins.
-
In the second sprint, it was decided to configure Next Social Login for application logins.
-
In the second sprint, it was decided to configure ACF + Custom Post Type for personalized content.
-
It was decided that the application would be at least 50% complete by the end of the second sprint.
-
### 📊 System and Application Architecture – Visual Descriptions
 |
 |
 |
|---|---|---|
| System Algorithm: It shows the entire process flow from the user's login to receiving recommendations. | Database Operations 1: It shows the structure where user information and test results are kept in the database. | Database Operations 2:It details role-based access and data backup processes. |
 |
 |
 |
| Web Application 1 | Web Application 2 | Web Application 3 |
The graph above is an up-to-date and comprehensive burndown chart based on Sprint 2's total of 200 points.
Red dashed line (Ideal Burndown): This is based on the assumption of equal progress each day.
Blue line (Actual Burndown): This shows the actual task completion rate.
Light blue area: This highlights the difference between ideal and actual burndown and is used to analyze time management.
- Sprint Notes:
• Database installation, backup, and role settings for professions and skills have been completed.
• A sample career recommendation model has been developed.
• Personality analysis tests have been selected for the NLP scorer, and integration has been planned.
• Artificial intelligence libraries (Pandas, NumPy) have been installed.
• The FastAPI environment for the backend has been established, and dependency management has been completed.
• Custom JavaScript, HTML, and CSS integrations have been implemented in WordPress.
• The decision to search the database using embeds has been made.
• Advanced security settings have been implemented with Google reCAPTCHA and WP 2FA.
• Social media login has been enabled with Nextend Social Login.
• Personalized content display has been configured with ACF + Custom Post Type.
• The overall functionality of the application is 55% complete.
- Estimated Score to Complete: 200 Points
- Prediction Logic
The scoring of tasks in Sprint 2 was determined by considering technical difficulty, effort intensity, and project impact. Tasks were scored as follows, in accordance with Google Project Management standards. Below is the scoring of tasks completed in Sprint 2, based on technical difficulty and project contribution:
| Quest Description | Point |
|---|---|
| Database setup, backup, and role settings for professions and skills | 20 |
| Developing a sample career recommendation model (AI model basis) | 25 |
| Personality test selection and integration plan for the NLP scorer | 20 |
| Installation of artificial intelligence libraries (Pandas, NumPy) | 10 |
| FastAPI backend installation and dependency management | 25 |
| Custom JavaScript, HTML, and CSS integration on WordPress | 15 |
| Implementation of semantic search functionality on the database with embedding | 15 |
| Security configuration with Google reCAPTCHA and WP 2FA | 15 |
| Nextend Social Login integration (social media logins) | 10 |
| Setup of personalized content structure with ACF + Custom Post Type | 15 |
| 55% completion of the application and implementation of the core infrastructure | 30 |
| Total Points: 200 |
This distribution shows that balanced progress was made in Sprint 2 in terms of both consolidating the software infrastructure and implementing AI and data-driven modules. The target score for Sprint 2 has been successfully achieved.
- Daily Scrum:
• Regular meetings were held via Google Meet at regular intervals.
• Continuous and instant communication was maintained via the WhatsApp group.
• Daily task tracking was maintained via Miro.
- Product Backlog URL: https://miro.com/app/board/uXjVIicQLWg=/?share_link_id=476687804339 (Miro)
- Sprint Review:
• FastAPI backend installation has been successfully completed.
• Database installation and user roles have been created, preparing for data management.
• Database search has begun with the initial embending for the career recommendation system.
• Login systems, user management, and security settings are being finalized.
• Test selection for personality analysis has been completed, and NLP integration has been planned.
• The application is 55% complete. UI details and final testing have been postponed to Sprint 3.
-
Sprint Review Participants: Tuba İlhan Horuz, Duygu Başak Acar, Sude Nur Kömür, Melih Talha Bektaş, Saim Berk Akçeşme
-
Sprint Retrospective:
• The coding process took longer than expected due to the intensity of the technical tasks.
• Creating a prediction model was abandoned.
• The decision was made to use database searches with embedding to produce more flexible, scalable, and dynamic results based on user input.
• Berk's work on the website increased focus and productivity.
• Melih and Duygu's focus on the AI component increased productivity.
• Task allocation via Miro facilitated task tracking.
• Communication within the team was strong, which helped us accelerate towards the end of the sprint.
• Most of the planned tasks were completed in the second sprint.
• Sprint 3 will focus on user testing, finalizing the NLP scorer, and finalizing UI/UX details.
- Other Notes:
The blue dashed line represents the ideal estimated burndown trajectory—completing 20 points per day. The green solid line shows the actual remaining work. As seen, the team slightly lagged behind the ideal pace during the first few days but gradually caught up. By Day 10, all 200 points were successfully completed. This reflects a well-managed sprint with effective adaptation and recovery.
-
Sprint Notes:
•Datasets are fully organized and ready for use.
•AI model development completed.
•The website has been finalized in line with the necessary needs and modernization requirements.
•Chatbot was made ready with the organized data sets and necessary operations.
•Necessary web integration was provided and the chatbot was connected.
•The website with artificial intelligence integration was ready for practical use after real tests.
-
Expected point completion within Sprint: 200 Points
-
Point Completion Logic: All project tasks in this sprint have been successfully completed. The point distribution was carried out in alignment with the Google Project Management methodology, taking into account technical complexity, time investment, and value contribution of each deliverable.
| Task Description | Points |
|---|---|
| Website finalized | 40 |
| AI model developed | 40 |
| AI–Web integration completed | 30 |
| Chatbot created | 30 |
| Datasets organized | 30 |
| Chatbot connection established | 30 |
| Total | 200 |
The sprint goal was to complete all planned technical tasks and deliver a fully functioning AI-integrated web platform. This goal was successfully achieved with a total score of 200/200.
The point-based breakdown ensured effective prioritization and resource allocation across the sprint. Each completed task directly contributed to the overall functionality and usability of the final product.
-
Daily Scrum:
•Since the delivery date was approaching and the product was required to be of much higher quality, some days were spent in online meetings.
• Regular meetings were held via Google Meet at regular intervals.
• Continuous and instant communication was maintained via the WhatsApp group.
• Daily task tracking was maintained via Miro.
-
Product Backlog URL: https://miro.com/app/board/uXjVIicQLWg=/?share_link_id=476687804339 (Miro)
-
Sprint Review:
The goal of this sprint was to finalize the AI-powered career guidance website by integrating all components, ensuring functionality through real-user testing, and making the system ready for practical use.
Completed Deliverables:
-
Datasets Organized: All datasets were cleaned, structured, and categorized based on project requirements.
-
AI Model Developed: A model was trained and validated to provide career recommendations based on user input.
-
Website Finalized: The user interface was updated and modernized to ensure a seamless user experience.
-
Chatbot Prepared: The AI model was integrated into a chatbot infrastructure to provide dynamic, data-driven responses.
-
Web Integration Completed: The chatbot was successfully embedded into the website, ensuring compatibility and functionality.
-
Real User Testing Conducted: The system was tested by users and proven to be fully functional, ready for deployment.
All planned items were successfully completed. With this sprint, the project reached a Minimum Viable Product (MVP) stage and is now ready for real-world use.
- Sprint Review Participants:
Tuba İlhan Horuz,Duygu Başak Acar,Melih Talha Bektaş,Saim Berk Akçeşme - Sprint Retrospective:
All goals were completed: The planned workload of 200 points was fully achieved by the end of the sprint.
Team communication was strong: Regular communication via WhatsApp and Google Meet enhanced team collaboration.
Data and model integration was successful: The AI model and chatbot system were successfully integrated.
Real user testing was conducted: The system was made ready for use and successfully passed testing.
- Other Notes:
- The project was developed using the Replit platform. Replit's online development environment facilitated seamless artificial intelligence integrations.
Technologies Used for Chatbot:
fastapi>=0.115.0 uvicorn[standard]>=0.32.0 pydantic>=2.10.0 pydantic[email]>=2.10.0
sqlalchemy>=2.0.36 alembic>=1.14.0
For PostgreSQL on Windows, use: pip install psycopg2-binary --no-build-isolation
google-generativeai>=0.8.3
pandas>=2.3.1 numpy>=2.1.0 python-multipart>=0.0.12
httpx>=0.28.1 aiohttp>=3.12.0 requests>=2.32.3
python-dotenv>=1.0.1 pydantic-settings>=2.6.1
passlib[bcrypt]>=1.7.4
psutil>=6.1.0
pytest>=8.3.4 pytest-asyncio>=0.24.0 pytest-cov>=6.0.0 black>=24.10.0 isort>=5.13.2 flake8>=7.1.1
click>=8.1.7 typer>=0.15.1
python-dateutil>=2.9.0
orjson>=3.10.12
Data-Driven and Responsive Routing
Career Path not only guides individuals with a data-driven approach, but also aims to raise awareness of the workforce impacts of the cycle of overproduction and overconsumption. It highlights the new career paths emerging from digital and green transformation, while also raising awareness of the long-term ecosystem impacts of career decisions.
An Impact-Focused Approach
As a platform, our goal isn't just to answer the question, "Which career should I choose?" but also to encourage individuals to ask themselves, "What impact do I want to leave?" Because we believe the workforce of the future should be comprised of individuals who are not only productive but also responsible and make a difference.
Goals for a Sustainable Future
One of Career Path's core goals is to reduce digital inequalities, promote climate-sensitive careers, and provide individuals with meaningful, productive, and sustainable career paths. Toward these goals, we leverage a robust technological infrastructure, a reliable data foundation, and a learning engineering approach that prioritizes student development.
Exploring the Professions of the Future
All of our efforts are aimed at empowering individuals to discover careers that will remain relevant in the future. This way, each individual can embark on a career journey that is more meaningful and sustainable for both themselves and the world.
- Needs Analysis: Students choose a career based solely on environmental factors.
- Data Sources: O*NET, TÜİK, İŞKUR, LinkedIn, Coursera, OECD
- AI-Powered Matching: Personality test results + market data
- Simulation: Virtual career experience
- Decision Support: Career recommendation percentages with AI model
| Technology | Purpose of Use |
|---|---|
| Python | Data analysis and modeling |
| Power BI | Profession data analysis and visualization |
| Coursera Trends / LinkedIn Insights | Big data analysis of skills |
| Canva | Student career plan presentations |
| Peplit | Website |
- Sectoral growth map
- Annual change in in-demand professions
- Student personal fit score and career prediction graph
Data-Driven Career Discovery/
│
├── data/ # Raw data and sample CSV files
├── dashboards/ # Power BI .pbix files and screenshots
├── models/ # AI model files (optional)
├── src/ # Python codes (analysis, matching)
├── screenshots/ # App screenshots
├── wordpress/ # Application website
├── README.md # Project introduction (this file)
ai_models/
│
├── base_ai.py/ # Basic AI class
├── scenario_ai.py/ # Scenario production
├── evaluation_ai.py/ # User rating
├── enhanced_career_recom...py/ # Advanced career advice
└── enhanced_evaluation_ai.py/ # Advanced evaluation
app/
├── core/ # Basic system components
│ ├── cache.py # Redis/Memory cache
│ ├── metrics.py # Performance metrics
│ └── security.py # Auth and security
├── database/ # Database layer
│ ├── connection.py # DB connection management
│ └── models.py # SQLAlchemy models
├── routers/ # API endpoints
│ ├── careers.py # Career APIs
│ ├── chat.py # Chat system
│ ├── evaluation.py # Evaluation APIs
│ ├── health.py # Health checks
│ ├── scenarios.py # Script production
│ └── users.py # User management
├── services/ # Business logic
│ ├── onet_data_processor.py # O*NET data processing
│ ├── user_profile_manager.py # Profile management
│ └── health.py # Health services
└── main.py # FastAPI implementation
├── app/ # Main application code
├── data/ # O*NET CSV data
├── career_env/ # Python virtual environment
├── ai_models/ # AI service modules
├── tests/ # Test files
├── logs/ # Log files
├── .github/workflows/ # CI/CD pipeline
├── .vscode/ # VS Code configuration
└── config files # Docker, Nginx, DB, etc.
We welcome pull requests and issues. To contribute:
1. Log in to our website now!
2. Take the quiz!
3. Share your results on social media!
4. Don't forget to tag us!!! 😊