Saaram AI - An AI Powered Study Companion with Speech Input and Speech Output

DOI

10.2991/978-94-6239-693-7_85How to use a DOI?

Keywords

Artificial Intelligence; Speech Recognition; Text-to Speech; Tanglish; Flutter; Natural Language Processing; Smart Study Companion; LangChain; Whisper

Abstract

Artificial Intelligence (AI) is now a part of the education system. Artificial Intelligence (AI) changes how students work with study materials and learn things. I have observed that speech recognition, natural language processing (NLP) and text-, to-speech synthesis have advanced quickly. Speech recognition, natural language processing (NLP) and text-to-speech synthesis push learning away from reading methods toward interactive and conversational experiences. This paper presents Saaram AI. Saaram AI works as the study companion that uses speech input and speech output. Saaram AI helps students understand materials, in an personalized way. Saaram AI lets you upload the PDF documents such, as lecture notes, research papers or textbooks. You can then ask questions using speech in a mixed format also called Tanglish. Saaram AI processes the voice queries with an optimized Python backend that runs FastAPI, LangChain and ChromaDB. Saaram AI then gives you answers as speech output, in the Tanglish tone. Saaram AI works fast. I find Saaram AI easy to use. When I upload a PDF document I ask a question in Tanglish. Hear the answer spoken back in Tanglish. When I look at the system I see that the system uses Whisper for speech recognition. The system uses LangChain for understanding and, for search. The system uses FAISS for vector based retrieval. The system uses gTTS (Google Text-to Speech) for speech output. The frontend of the system is built with Flutter. Flutter makes the user interface simple. Works on platforms. Saaram AI talks in a tone that feels familiar to the region. Saaram AI creates a learning environment that's inclusive, relevant, to the region and engaging. I tested the application, with school papers. The application gave ninety percent speech recognition accuracy and between eighty-eight and ninety-one percent contextual relevance. I mixed speech input with language answers. Saaram AI shows a step, toward making AI learning companions that're smart and talk like people.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - S. A. Maneesha
AU  - S. Meiyammai
AU  - V. Gowri Manohari
PY  - 2026
DA  - 2026/06/16
TI  - Saaram AI - An AI Powered Study Companion with Speech Input and Speech Output
BT  - Proceedings of the International Conference on Intelligent Systems for a Sustainable Future (ISSF 2026)
PB  - Atlantis Press
SP  - 869
EP  - 881
SN  - 2589-4919
UR  - https://doi.org/10.2991/978-94-6239-693-7_85
DO  - 10.2991/978-94-6239-693-7_85
ID  - Maneesha2026
ER  -