Online Career Mentorship Platform Using AI-Powered Recommendations

Abstract

Traditional mentorship programs have long depended on manual pairing and face-to-face sessions, creating barriers related to availability, scale, and geographic reach. With the rapid expansion of digital education and remote work, there is growing demand for structured, technology-driven alternatives that can match students and professionals with the right mentors while also supporting ongoing career guidance. This paper presents an Online Career Mentorship Platform that employs a three-layer AI architecture: a hybrid recommendation engine combining collaborative filtering with BERT-based semantic profile matching, a large language model (LLM)-powered conversational assistant for real-time query handling and session management, and a sentiment-driven feedback loop for continuous platform improvement. The proposed system was evaluated on a dataset of 1,200 user profiles and 8,400 historical mentorship interaction logs. Matching accuracy reached 94.3%, chatbot response relevance scored 91.7%, and mentee satisfaction climbed to 88.6% compared to 61.2% under manual matching. The paper also examines the system\'s modular architecture, discusses practical deployment challenges including data privacy and cold-start problems, and charts the platform\'s trajectory toward multimodal and emotionally-aware mentorship.

Introduction

This text describes an AI-powered mentorship and career guidance platform designed to reduce inequality in access to career opportunities by providing intelligent mentor matching and conversational support.

The main problem identified is that career guidance today is unevenly distributed—students with strong networks get mentorship easily, while others rely on informal, passive systems like email requests or forums that often go unanswered. Existing platforms lack structured matching, personalization, and feedback-driven improvement.

To solve this, the proposed system combines three core AI components:

1. Semantic mentor matching using BERT embeddings,
2. Collaborative filtering based on past interactions, and
3. A conversational AI chatbot powered by LLMs and retrieval-augmented generation (RAG) for career guidance and administrative tasks.

Mentor and mentee profiles are converted into 768-dimensional embeddings using a fine-tuned BERT model, allowing semantic understanding of career goals. Matching is done using a hybrid scoring system that combines semantic similarity and collaborative filtering, with adaptive weighting to handle cold-start users. This ensures both new and experienced users receive relevant mentor recommendations.

The system also includes a chatbot assistant that handles scheduling, queries, and career advice using intent classification and RAG-based responses to ensure accuracy and reduce hallucinations. Additionally, it integrates calendar scheduling, reminders, and feedback collection, making the mentorship process more automated and structured.

Feedback is analyzed using sentiment analysis (DistilBERT-based) to improve future recommendations and monitor engagement. Low engagement triggers automated outreach to re-engage users.

In evaluation, the system was tested with 340 mentees and 95 mentors over six months. The hybrid model achieved 94.3% top-5 recommendation accuracy, outperforming both rule-based and purely semantic approaches. The chatbot also showed high intent recognition accuracy (~95%) and strong response quality.

Overall, the platform demonstrates how combining semantic AI, collaborative filtering, and conversational systems can create a scalable and intelligent mentorship ecosystem that improves accessibility, personalization, and engagement in career development.

Conclusion

This paper has described an online career mentorship platform that brings together three AI components — hybrid semantic and collaborative mentor matching, an LLM-backed conversational assistant, and a sentiment-aware feedback loop — into an integrated system designed to make quality mentorship more accessible and more scalable. The pilot evaluation demonstrated meaningful improvements over both manual matching and simpler algorithmic alternatives across the key metrics of matching accuracy, chatbot relevance, and mentee satisfaction. The platform\'s architecture is deliberately modular, so that individual components can be improved or replaced as AI capabilities continue to advance without requiring a full system redesign. This matters because the field is moving quickly: the transformer models, LLM APIs, and recommendation algorithms that represented the state of the art at the start of this project have already been substantially advanced, and the next several years are likely to bring further changes at a similar pace. What seems unlikely to change is the fundamental value of a well-matched mentorship relationship. The evidence that mentorship accelerates career development, improves job satisfaction, and expands professional networks is well-established [10]. The challenge has always been access — making sure that the benefit is not limited to those who happen to be in the right place at the right time. A platform that uses AI to close that gap, carefully and with appropriate attention to fairness and privacy, is working toward something worth building

References

[1] Devlin, J., Chang, M. W., Lee, K., and Toutanova, K. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. Proceedings of NAACL-HLT 2019, pp. 4171–4186. [2] Eby, L. T., McManus, S. E., Simon, S. A., and Russell, J. E. A. (2000). The protégé perspective regarding negative mentoring experiences: the development of a taxonomy. Journal of Vocational Behavior, Vol. 57, pp. 1–21. [3] Pazzani, M. J., and Billsus, D. (2007). Content-based recommendation systems. In The Adaptive Web, Lecture Notes in Computer Science, Vol. 4321, pp. 325–341. Springer, Berlin. [4] Zain, A., and Farooq, U. (2023). Student Support Chatbot Using BERT and Rule-Based NLP for Career Guidance. International Journal of Interactive Mobile Technologies, Vol. 17, No. 11, pp. 45–56. [5] Shilaskar, S., Agarwal, P., Joshi, A., and Kale, V. (2024). ASPIREAI: AI Chatbot for Career Guidance. Proceedings of the International Conference on Artificial Intelligence and Applications, pp. 112–118. [6] Singh, R., and Rao, A. (2024). ChatGPT and Career Counselling: A New Paradigm in Student Support Systems. International Journal of Innovative Research in Computer and Communication Engineering, Vol. 12, No. 4, pp. 2449–2454. [7] Lade, S., Billade, A., Chandrapatle, S., Chenna, S., and Chinchalpalle, G. (2024). CareerCue: A Survey on Intelligent Job Recommendation Systems Leveraging AI and ML. International Journal of Scientific Development and Research, Vol. 10, No. 8, pp. 1823–1831. [8] Pourheidari, V., Sotoodeh Mollashahi, E., Vassileva, J., and Deters, R. (2024). Recommender System based on Extracted Candidate and Job Features Using Graph Neural Networks. Proceedings of the IEEE International Conference on Data Mining Workshops, pp. 340–347. [9] Panda, S., and Mehta, A. (2024). AI-Based Multilingual Career Assistant for Rural Students. International Journal of Educational Technology and Learning, Vol. 12, No. 3, pp. 112–119. [10] Dey, S., and Bose, T. (2023). Conversational AI in Student Mentorship: Opportunities and Challenges. Asian Journal of Computer Science and Information Technology, Vol. 13, No. 1, pp. 15–21. [11] Fagenon-Eland, E. A., Marks, M. A., and Amendola, K. L. (1997). Perceptions of mentoring relationships. Journal of Vocational Behavior, Vol. 51, pp. 29–42. [12] Guan, J., and Li, M. (2024). Skill-Aware Job Recommendation via Semantic-Enhanced Transformers. Proceedings of ACL 2024, pp. 888–900. [13] Bisaso, S., Wasswa, W., and Muhumuza, G. (2025). Towards the Application of an NLP-Driven Conversational Model for Efficient and Affective Career Guidance for Students: A Scoping Review. European Journal of Applied Science, Engineering and Technology, Vol. 3, No. 3, pp. 240–253. [14] Bagai, R., and Mane, V. (2024). Designing an AI-Powered Mentorship Platform for Professional Development: Opportunities and Challenges. arXiv preprint arXiv:2407.20233.

Copyright

Copyright © 2026 Rahul Bhardwaj, Dr. Sanjay Pachauri . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.