TEXTOIR is the first opensource toolkit for text open intent recognition. (ACL 2021)

[RA-L21] Long-Term Pedestrian Trajectory Prediction Using Mutable Intention Filter and Warp LSTM

Lightweight, intent-driven control plane for smart home & local automation. Uses iOS Shortcuts & Control URLs without local LLM/GPU hosting.

Python library to extract intention and structured information from sentences written in natural language

ChatAssist Pro is an advanced AI-powered chatbot designed to streamline customer support interactions.

Interactive NLP-based AI system designed to manage cinema bookings and provide a seamless user experience.

Research on the Construction Method of Intelligent Services Based on Prior Knowledge

Python NLP chatbot for train ticket booking and journey planning, featuring intent recognition, dialogue management, and evaluation.

A serverless, event-driven chatbot architecture using Amazon Lex, AWS Lambda, and DynamoDB for scalable intent recognition and dynamic user interaction.

Full-Stack, NLP, Voice AI | 24/7 Automated Customer Support

Constructing a corpus of ancient Chinese pediatric medicine literature, using algorithms such as BERT, lattice LSTM, and Siamese for tasks such as named entity recognition, intent recognition, entity similarity calculation, and entity linking, to develop a TCM-QA.

Simple voice assistant made for use in Russian

Production-style intent recognition system: transformer embeddings, multi-model inference (LR/SVC), offline STT, dynamic skill modules, and a full Streamlit demo UI. Fast, modular, and extensible. IN SHORT : Modular intent recognition engine with transformers, ML models, and real-time inference.

Fully Local AI Assistant

Practical from my Social Signal Processing class

Built a secure HR chatbot that lets employees ask questions like leave balance, salary details, and other personal info. Each employee can only see their own data, making it a safe and easy way to access HR services.

Bachelor thesis regarding in-domain intent recognition and out-of-domain detection.

Successfully developed a Multi-Source Knowledge Intelligence Agent - an advanced agentic AI application that integrates multi-format document and web data processing with LLM-powered analytics and conversational intelligence. It analyzes multi-source data (URL/files) with AI-driven insights, sentiment analysis, knowledge graphs, and chat interface.

A powerfull NLP-Based application that analyzes email content and generates optimized subject lines with tone(like an overview)