Advanced Search
Search Results
7 total results found
Worker
The Worker is a background daemon that runs alongside the CRM. When a task takes a long time (such as chatting with an AI, running a process, uploading documents for a vector search, etc..), the Worker handles it asynchronously so the browser does not freeze a...
Orchestrator operations
Requirements llama-cpp-python is installed as a pre-built wheel (not compiled from source). requirements.txt specifies the Vulkan variant via --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/vulkan. Other backends are available by changing th...
Install docker for Kitt
Ubuntu #!/bin/bash set -euo pipefail SELF="${BASH_SOURCE[0]}" SELF="$(realpath -ms "$SELF")" cd "$(dirname "$SELF")" || exit 2 if ((UID)) && ! [[ -v SUDO_USER ]]; then exec sudo "$SELF" "$@" fi curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /et...
Worker architecture
Default setup flows The Worker is composed of three processes that work together: 1. Router File: include/Services/Worker/Router.php The Router listens on a Unix socket for incoming connections. When a browser tab or a PHP script connects, the Router identifi...
Worker operations
Installation sudo tools/worker install This copies the systemd unit files ( vte-worker@.service and vte-worker@.socket) to /etc/systemd/system/, enables the socket, and starts it. Both files are systemd templates that take the relative path from /var/www/ht...
Orchestrator architecture
Three layers: PHP CRM (Worker Consumer), calls orchestrator via HTTP Python orchestrator (FastAPI, LangChain) LLM / MCP servers / Chroma Relevant Worker Consumer methods: llmChat() (chat relay), elaborateRag() (document upload + vector build). RAG Indexing T...
Agent orchestrator
Python service providing AI agent chat, MCP tools integration, and RAG on CRM documents, based on FastAPI and LangChain. Called by the Worker's Consumer processes via HTTP/SSE. See the dedicated chapter here.