An AI-Driven Training and Placement Platform with Predictive Analytics and Conversational Assistance

Read full text Summarise with VIRAT
Authors: Srikar G Kulkarni, Vaishnavi Kamthe, Kumar Saransh, Nemat Momin, Sonali Shirke, Mukul Jagtap
Cited by: 0
References: 31

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The increasing volume and diversity of student performance data have exposed significant limitations in traditional training and placement systems, which primarily rely on static eligibility criteria, manual shortlisting processes, and delayed communication mechanisms. These systems also lack the capability to leverage predictive analytics, resulting in minimal personalized, data-driven insights to enhance student employability based on individual skills and qualifications. Although several studies have applied machine learning techniques for job applicant ranking, most existing solutions lack real-time integration, interpretability, and conversational support within placement systems. To address these challenges, this study proposes an AI-driven education and placement platform that integrates machine learning-based placement prediction with conversational assistance and intelligent job matching. The system utilizes XGBoost for predictive modeling, Sentence-BERT embeddings for semantic skill representation, SHAP for explainable insights, and Retrieval-Augmented Generation (RAG)-based chatbots to provide real-time guidance and interview preparation support. The platform is implemented using FastAPI and deployed on cloud infrastructure, with automated email notification systems enabling real-time user interaction. The proposed system was evaluated using a dataset of 1,200 student records, incorporating academic, skill-based, and experiential attributes. Experimental results demonstrate an accuracy ranging from 88% to 90%, along with strong performance across multiple evaluation metrics, including precision, recall, F1-score, and ROC-AUC. Additionally, the system achieved low inference latency (

Fields of Study & Subjects

AI and HR Technologies Online Learning and Analytics Artificial Intelligence in Education

Publication details

Journal / source
Journal of Smart Sensors and Computing
Volume
2
Issue
1
Publication date
30 March 2026
Language
EN