Transform your documents into intelligent conversations. This open-source RAG chatbot combines semantic search with fine-tuned language models (LLaMA, Qwen2.5VL-3B) to deliver accurate, context-aware responses from your own knowledge base. Join our community!

It's a cache for text embeddings, but intelligent

Semantic Document Processor is an intelligent, open-source document analysis system that leverages Large Language Models (LLMs) to process and analyze complex documents like insurance policies, contracts, and legal documents. Built with modern AI/ML technologies, it provides structured, decision-ready responses to natural language queries

Diffusion-guided topic modeling using semantic embeddings and unsupervised learning. Combines Sentence-BERT, diffusion models, and clustering to discover coherent topics, generate topic-conditioned text, visualize topic structures, and compare with traditional methods.

🛡️ An AI-powered debate training and communication risk analysis platform. LogicShield strengthens your arguments, detects logical fallacies, and evaluates reputational risk before you publish, pitch, or perform.

Natural language processing project combining semantic embeddings and toxicity detection to analyze Brexit-related immigration discourse.

Physics-based clustering of embedding vectors using N-body gravitational diffusion on the unit hypersphere. Anisotropic per-dimension forces, adaptive-threshold connected-components extraction. Julia + KernelAbstractions, AMDGPU + CPU backends

Game recommendation system leveraging the all-MiniLM-L6-v2 sentence transformer for embedding generation and FAISS index for efficient similarity-based retrieval.

Cognitive MRI of AI Conversations: Network analysis of ChatGPT conversation logs using semantic embeddings to reveal knowledge topology, community structure, and cross-domain bridges

Hybrid multi-document summarization system using Sentence-Transformers and KMeans clustering for semantic topic extraction, followed by BART-based abstractive refinement. Fully containerized with FastAPI and Docker.

Interactive 3D codebase visualization using AST parsing, semantic embeddings, and graph analysis. Built with WebAssembly, WebGPU, and modern web APIs.

Automated vertical short-form video assembly that matches each script sentence to stock footage via semantic embeddings, then stitches clips, voiceover, and styled subtitles into a finished video.

AI-powered resume screening and ranking system using semantic embeddings and multi-factor scoring for fair and scalable recruitment.

Probabilistic day-ahead LMP forecasting for 8–12 PJM zones using Temporal Fusion Transformer and a frozen semantic-embedding fusion branch. Produces P10/P50/P90 hourly forecasts, attention and variable-importance visualizations, and fusion-gate diagnostics.

📊 Research-driven NLP pipeline for labor-market analysis, using sentence embeddings, clustering, and taxonomy alignment with O*NET / SOC codes.

Experimental computational framework for semantic exploration of the Voynich Manuscript using transformer embeddings, medieval corpora, semantic clustering and digital humanities methodologies.

Official implementation of the paper: Papazis, S.; Giotis, A.P.; Nikou, C. "Enhancing Keyword Spotting via NLP-Based Re-Ranking: Leveraging Semantic Relevance Feedback in the Handwritten Domain." Electronics 2025, 14, 2900.