research / PROJECT NOTE / 2026
SENTIENT: a local assistant for text, images, PDFs and audio
A modular multimodal assistant using LLaMA 2, LLaVA and Whisper.
4input types in one local assistant
Python · LangChain · Ollama · Streamlit · LLaMA 2 · LLaVA · Whisper
SENTIENT is a locally deployed conversational multimodal AI assistant for text, images, PDFs and audio. I integrated LLaMA 2, LLaVA and Whisper using Python, LangChain, Ollama and Streamlit.
A modular LangChain pipeline orchestrates the models for context-aware interaction and local inference. The project brought language, vision and speech processing into one conversational interface.
[Explore my GitHub]Project overview adapted from Shishir Mishra’s supplied resume.
