L

LLMs

LLM Workflow Router is a stateless middleware engine designed to enforce explicit execution topology in AI systems that rely on large language models. It evaluates interaction metadata against strictly declared workflow rules and returns a terminal state.

A unified Python interface to select and use multiple Large Language Model (LLM) providers through a common API.

💬 Epic prompts to turbo-charge your LLM chatbots.

a self-hosted, locally-routed LLM stack with a default-deny security posture. your models. your machine. your terms.

AI Image Renamer is a Python CLI tool that leverages generative artificial intelligence to rename your images based on what the image shows.

KiM Explorer is a two-stage RAG application for transport policy research publications from the KiM Netherlands Institute for Transport Policy Analysis. Users perform semantic search to identify relevant documents, manually select publications, then interact with an LLM using full document context rather than chunks. Built with Python/NiceGUI/OpenAI API, featuring citation generation, conversation history, filtering, and web/CLI interfaces. https://explorer.kim.rijkscloud.nl/

AI Governance Control Platform

Ce projet consiste à concevoir un assistant juridique intelligent permettant à un utilisateur d'interagir en langage naturel avec un corpus documentaire basé sur le Code pénal sénégalais.

L'objectif est de mettre en œuvre une architecture Retrieval-Augmented Generation (RAG) afin de permettre au modèle de langage de s'appuyer sur des informations issues du corpus documentaire avant de générer une réponse.

Le projet explore ainsi l'utilisation des LLM, des embeddings et de la recherche sémantique pour construire une application capable de transformer des documents juridiques en une base de connaissances interrogeable.

Travaux réalisés :

Collecte et préparation du corpus juridique. Nettoyage et structuration des documents. Découpage des documents en unités adaptées à la recherche. Transformation des passages textuels en représentations vectorielles (embeddings). Mise en place d'un mécanisme de recherche sémantique. Récupération des passages les plus pertinents en fonction de la question utilisateur. Intégration des informations récupérées dans le contexte transmis au LLM. Génération de réponses contextualisées à partir des documents récupérés. Conception d'une interface permettant d'interagir avec l'assistant en langage naturel.

Kevlar Benchmark: OWASP Top 10 for Agentic Apps (AI-Agents) 2026 a Red Team Benchmark.

Paper: "A Comprehensive Evaluation of Pre-Trained Language Models for Irony Detection in Tweets"

Red Team AI Benchmark: Evaluating LLMs for authorized offensive-security tasks. Red Team AI Benchmark is a CLI model-evaluation benchmark. It measures how LLMs understand and respond to red-team questions and security scenarios; it is not a tool for carrying out those activities. Version 2 uses a rubric-based dataset instead of judging answers only against one golden response.

This project is a Retrieval-Augmented Generation (RAG) application built using LangChain. It leverages advanced language models and vector databases to answer questions about epidemiological modeling, software development, and maintaining the EPP model for HIV modeling.

Production-ready RAG starter: hybrid search, chunking strategies, observability (Prometheus/Grafana), MLflow tracking, drift detection, GDPR deletion, and evaluation. The parts the tutorials skip.

Multimodal architecture that is capable to sensing emotion from chat occurrences, with added support for emojis, emoticons, and internet slangs.

Flexible GraphRAG: Python, LlamaIndex (LangChain also coming) Docker Compose: 8 Property Graph dbs, 3 RDF graph dbs, 10 Vector dbs, OpenSearch, Elasticsearch, Alfresco. 13 data sources (9 auto-sync), KG auto-building, RDF ontologies, schemas, LLMs, document processing, Docling, LlamaParse, GraphRAG, RAG only, Hybrid search, AI chat. React, Vue, Angular TypeScript frontends, FastAPI backend, REST API, MCP Server

A minimal Python3 program to collect and store JSON responses concurrently from LLM models using zai-sdk.

Deterministic Reasoning Framework (DRF)

A framework for Trustworthy AI in Critical Decision Making

The Deterministic Reasoning Framework (DRF) is a command-line experimental tool designed to demonstrate a fundamental thesis in AI safety and reliability: critical decision-making tasks involving classification or ranking should not rely on pure LLM reasoning due to its stochastic and non-reproducible nature.

NexusMind is a project centered around two Docker containers, llm_proxy and llm_redis. These containers are designed to act as your personal proxy to your favorite LLM, logging all inbound and outbound tokens to Redis for future use.

🤖 nGPT: A Swiss army knife for LLMs: powerful CLI and interactive chatbot in one package. Seamlessly work with OpenAI, Ollama, Groq, Claude, Gemini, or any OpenAI-compatible API to generate code, craft git commits, rewrite text, and execute shell commands. Fast, lightweight, and designed for both casual users and developers.

https://nazdridoy.github.io/ngpt/