MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202621031885 A) filed by Prof. Neha Prabhakar Rajas; Prof. Sheela Vijay Chinchmalatpure; Prof. Surbhi Sanjiv Kakde; Atharva Devendra Kavade; Ayushi Mahendra Kamble; Srushti Popat Kasurde; Sushant Anil Katare; and Vivek Namdevrao Kendre, Pune, Maharashtra, on March 17, for 'ai interview coach.'

Inventor(s) include Prof. Neha Prabhakar Rajas; Prof. Sheela Vijay Chinchmalatpure; Prof. Surbhi Sanjiv Kakde; Atharva Devendra Kavade; Ayushi Mahendra Kamble; Srushti Popat Kasurde; Sushant Anil Katare; and Vivek Namdevrao Kendre.

The application for the patent was published on May 29, under issue no. 22/2026.

According to the abstract released by the Intellectual Property India: "An AI-powered virtual interview coaching system that is computer-implemented has been disclosed which offers adaptive and personalized interview preparation in both technical and HR interview domains. The system uses transformer-based language models for natural language processing: Flan-T5 to generate questions, MiniLM for measuring the semantic similarity of answers, Mistral-7B for writing personalized feedback, and Whisper for converting speech to transcription. The whole setup is able to create real-life interview experiences without a need for specialized hardware. The system compares user replies through a multi-level evaluation pipeline integrating semantic similarity analysis, keyword and concept matching, grammar checking, knowledge graph-based error identification, sentiment, and confidence estimation, resulting in a combined score and meaningful personalized feedback. A model that determines the level of difficulty in an adaptive way changes the complexity of questions based on the rolling average performance data, so that learners remain on the ideal learning path. The system currently supports domains such as Artificial Intelligence, Machine Learning, Data Structures and Algorithms, Full Stack Development, Operating Systems Java Human Resources interviewing and it is aimed to be accessible by students, job seekers, and working professionals via regular computing devices."

Disclaimer: Curated by HT Syndication.