Introduction
LearnerOn.Net was an innovative project based in Prague, CZ, focused on developing a learning experience platform specifically designed for corporate employees. As the AI Model Engineer and tech lead, I played a pivotal role in the development and implementation of custom graph neural networks (GNNs) to map and understand the complex relationships between various skills, job positions, and learning resources. The primary objective was to create personalized learning paths that would empower employees to advance their careers by acquiring the skills needed for their desired job positions. This platform aimed to enhance employee development and retention by offering tailored learning recommendations that aligned with individual career goals.
Target Group
- Primary: Corporate Employees seeking career advancement and skill development.
- Secondary: Corporate Learning and Development (L&D) Departments, HR Teams, and Talent Management Professionals.
Project Objectives
- Personalized Learning: To deliver personalized course and learning material recommendations that align with each employee's current skillset and career aspirations.
- Career Advancement: To provide employees with clear, actionable learning paths that guide them toward their target job positions within the company.
- Skill Gap Analysis: To enable employees to compare their current skills with those required for desired positions, identifying areas for improvement and guiding learning efforts.
What We Did for the Client
- Graph-Based Recommendation Development
- Scope of Work: I engineered a custom graph-based recommender system using graph neural networks (GNNs). This system was designed to generate embeddings for courses, skills, and job positions, which were then used to understand and analyze the relationships between these elements.
- Process:
- Data Collection and Structuring: We gathered and structured data on employee skills, job positions, and available learning resources. This data was then organized into a graph database, where nodes represented skills, courses, and job positions, and edges represented the relationships between them.
- Embedding Creation: Using GNNs, we created embeddings that captured the semantic relationships between different skills, job positions, and learning resources. These embeddings were crucial for making accurate and personalized recommendations.
- Recommendation Generation: The system analyzed the graph to generate personalized recommendations for each employee, suggesting courses and learning materials that would help them close skill gaps and advance toward their career goals.
- Skillset and Job Position Comparison
- Functionality: The platform allowed employees to input their current skills and desired job positions. It then compared these skills with those required for the target position, highlighting gaps and suggesting learning paths to bridge these gaps.
- Dynamic Learning Paths: Based on the comparison, the platform generated dynamic learning paths that recommended specific courses and learning resources. These paths were tailored to the employee’s current skill level and career objectives, providing a clear roadmap for professional development.
- Database Model and Ontology Structure
- Database Design: I assisted in defining the database model and ontology structure, ensuring that the data was efficiently organized for fast retrieval and analysis. The ontology captured the hierarchical relationships between skills, job positions, and learning resources, enabling the platform to deliver accurate and context-aware recommendations.
- Data Integration: The database was integrated with various internal and external data sources, including HR databases, learning management systems (LMS), and external course providers. This integration ensured that the platform had access to the most relevant and up-to-date learning materials.
- Research Code Productionization
- Transitioning to Production: I led the transition of research code into production-ready applications. This involved optimizing the code for stability, scalability, and performance, ensuring that the platform could handle large volumes of data and deliver recommendations in real-time.
- Scalability and Performance: The productionization process included implementing scalable architecture to support the needs of large corporations, where thousands of employees could use the platform simultaneously.
- Trend Monitoring and Innovation
- Staying Ahead of Trends: To ensure the platform remained cutting-edge, I continuously monitored industry trends and technological advancements in AI, particularly in the fields of NLP, machine learning, and graph theory.
- Incorporating Innovations: New findings and technologies were regularly incorporated into the platform, ensuring that it leveraged the latest innovations to provide the best possible recommendations and user experience.
Key Challenges
- Complex Relationship Mapping: Accurately mapping and understanding the complex relationships between skills, job positions, and learning resources required sophisticated graph neural networks and precise data modeling.
- Personalization at Scale: Delivering highly personalized recommendations to potentially thousands of employees in large corporations posed significant technical challenges, particularly in terms of scalability and performance.
- Data Integration and Consistency: Integrating data from various sources and ensuring its consistency across the platform was crucial for delivering accurate and reliable recommendations.
Outcomes and Client Impact
- Enhanced Employee Development: The platform successfully provided employees with personalized learning paths, helping them acquire the skills necessary for career advancement within their companies. This led to higher employee satisfaction and improved retention rates.
- Improved L&D Efficiency: By automating the process of skill gap analysis and learning path generation, the platform significantly improved the efficiency of corporate Learning and Development (L&D) departments.
- Strategic Talent Management: The platform enabled HR teams to better manage talent development, ensuring that employees were continuously developing skills that aligned with the company’s strategic goals.
Conclusion
LearnerOn.Net was a transformative project that redefined corporate learning and development through the use of advanced AI and graph neural networks. By delivering personalized learning recommendations and dynamic career advancement paths, the platform empowered employees to take control of their professional growth. It also provided corporate L&D departments with powerful tools to manage and optimize employee development on a large scale. This project demonstrated the potential of AI-driven platforms to enhance talent management and employee satisfaction, making LearnerOn.Net a valuable asset for modern corporations.
