Picture this: You have an AI-powered assistant that helps manage your team’s growth. It works with three types of text data: job descriptions, employee skills, and training programs. You give it a job description and an employee’s skill set, and it pinpoints the perfect training programs to bridge any skill gaps. It’s like having a tailor-made career coach that ensures your team is always ready for their next big challenge.
We have developed precisely this kind of solution — an AI-based text classification and recommendation tool that intelligently maps and matches textual data across extensive knowledge databases. By leveraging advanced Al algorithms, this solution can analyze and categorize vast amounts of text, providing relevant recommendations and insights.
Our solution has successfully passed the AWS Foundational Technical Review (FTR), a rigorous process conducted by AWS experts. The FTR ensures that solutions meet AWS’s best practices for security, reliability, and operational excellence, providing clients with a trusted, high-quality product that seamlessly integrates into the AWS ecosystem.
To better understand the capabilities and impact of our solution, let’s apply it to a specific area where it is currently in use — higher education institutions. We have customized this solution for our client who provides a largest Enterprise Resource Planning (ERP) platform to educational institutions in the United States.
Addressing challenges in higher education
Students often struggle to select the right courses and identify skill gaps, while administrators face challenges managing extensive course catalogs and maintaining up-to-date skills databases. Our AI solution addresses these issues by recommending personalized courses based on students’ profiles and career goals, ensuring they fill skill gaps without repetition. It also provides dynamic skill maps and timely course notifications. For administrators, it simplifies course management by automating skill extraction and updating, as well as generating engaging course descriptions, keeping the skill graph current and comprehensive.
Watch our AI solution in action
In this video, you will see how our AI solution enhances educational processes. We will demonstrate how our application updates existing course skills, creates new courses with optimal skill sets, provides personalized course recommendations for students, suggests additional career paths, and assists administrators in creating course content.
How it works
Our solution utilizes the Retrieval-Augmented Generation (RAG) approach, a reference architecture designed to enhance the functionality and performance of Large Language Models (LLMs) tailored for educational sector. The integration of the RAG method allows our solution to dynamically retrieve relevant context and enrich the LLM prompts with targeted data. Consequently, this approach eliminates the need for deploying and fine-tuning custom LLMs.
Key components of RAG architecture
Here are our system’s key components, unified by RAG principles:
User console: This is the primary interface where users (students or administrators) input their queries. The process begins with a user prompt and query.
Search module: The query is sent to the search module, which retrieves relevant information from the Neo4j database.
Neo4j database: The Neo4j database is the core of our knowledge sources. It stores and organizes the skill taxonomy, course data, and student profile data, allowing for efficient retrieval and updating of information.
Large language model (LLM): The user query, along with the prompt, is sent to the LLM (Claude3 from Anthropic) for processing. In this scenario, the model generates skill suggestions based on the query and the retrieved data from the Neo4j database. The output from the LLM is sent back to the User Console, providing the user with generated skill suggestions based on their initial query.
Data matching logic
The core logic of this solution, customized for higher education institutions, revolves around three key data categories mentioned earlier
- Skills taxonomy: This is a categorization system for skills, much like how taxonomy classifies the animal kingdom. Our solution leverages an existing taxonomy that categorizes and organizes various skills relevant to higher education.
- Course data: Each course in our system is associated with certain skills that students can acquire upon completion. This mapping allows the system to understand what skills are offered by which courses.
- Student profiles: Each student profile includes the skills a student has already acquired from completed courses and the career paths they have chosen. The profile also indicates which additional skills are needed based on these career paths.
The RAG approach helps the system “read” vast amounts of data on courses, skills, and student profiles to match students’ required skills with appropriate courses. Here’s how it works:
- Skills assessment: The system evaluates a student’s profile to determine which skills they already possess from completed courses.
- Career path analysis: Students select their desired career paths, each of which requires a specific set of skills. The system identifies the skills the student still needs to acquire based on their chosen paths.
- Course recommendation: The system then recommends courses that provide the missing skills, ensuring the recommendations are tailored to the student’s current skill set and career goals. It avoids suggesting courses that cover skills the student has already mastered.
This approach ensures students receive relevant course recommendations to progress along their desired career paths.
Get in touch to transform your organization with AI
Whether you’re in education, business, or any industry dealing with large text datasets, our solution can be tailored to meet your unique needs.
Don’t miss out on the opportunity to leverage AI for smarter content management and dynamic recommendations. Contact us today to learn more about how we can implement a customized AI solution for your organization. Let’s work together to unlock new possibilities and drive innovation.
Reach out to us now and start your journey towards an AI-powered future.
