lllm

Standards for building agents, better

Analyse environmental, social, and governance policies for potential gaps using AI - powered by LLMs, RAG techniques, and a regulatory knowledge base using OpenAI API, LangChain, Pinecone and Streamlit.

Plurals: A System for Guiding LLMs Via Simulated Social Ensembles

Typescript library for integrating Adyen APIs via an MCP server

Let's get voice feedback with claude code hooks

Ollama MCP Server, let's you chat with your local LLM running on your local OLLAMA server.

SVCA is a pipeline for automatically generating and uploading short-form video content to YouTube based on trending topics.

Fais - It does what you tell it to do

a multilingual tokenizer written from scratch in go

This project demonstrates how to setup a complete Retrieval Augmented Generation (RAG) pipeline on medical data using Llama-3-8B model

Open-source, self-hostable speech-to-speech AI voice platform. Real-time phone calls with sub-1.5s latency, built with Twilio, Deepgram, Gemini, and Cartesia. MIT licensed.

agent harness cli and api. "there are many like it, but this one is mine."

GAFA takes any text or image, extracts verifiable claims, and cross-references them against the Google Fact Check API and user-provided documents to deliver evidence-based verdicts in real time.

Autonomous 3D AI agent inspired by Caine from The Amazing Digital Circus, featuring perception, memory, and LLM-driven behavior.

以北宋熙宁变法为背景的 AI 原生历史决策模拟游戏。玩家亲拟诏书,由语言模型解析为政策并演绎其在官僚体系中的传导与失真。

Production-ready WhatsApp AI dealership assistant with grounded RAG, Google Calendar booking, APScheduler reminders, FSM orchestration, Hinglish support and anti-hallucination guardrails.

MCP server for Google Trends with proxy rotation and auto-retry on Google blocks connects agents to live keyword interest, related queries, and regional popularity.

A simple, minimalistic, and explainable code implementation of of Nemotron 3 Nano in JAX

AI-powered customer interview analysis tool — semantic search, insight extraction, and research synthesis built for the Great Question stack.

A theme park for your AI's pages — DESIGN.md themes that stop generic-looking output