Skip to content

gitlawr/appilot

 
 

Repository files navigation

Appilot

Appilot['æpaɪlət] stands for application-pilot. It is an experimental project that helps you operate applications using GPT-like LLMs.

Feature

  • Application management: deploy, upgrade, rollback, etc.
  • Environment management: clone, view topology, etc.
  • Diagnose: view logs, find flaws and provide fixes.
  • Safeguard: any action involving state changes requires human approval.
  • Hybrid infrastructure: works on kubernetes, VM, cloud, on-prem.
  • Multi language support: It's not restricted to a specific natural language. Choose the one you're comfortable with.
  • Pluggable backends: It supports multiple backends including Walrus and Kubernetes, and is extensible.

Demo

Chat to deploy llama-2 on AWS:

appilot-llama2.mov

Other use cases:

Run

prerequistes:

  • Get OpenAI API key with access to the gpt-4 model.
  • For Walrus backend
    • Install Walrus and get the url and API key. Walrus is an open source software that can be run by a docker command. It serves as the engine for application management.
  • For Kubernetes backend
    • Install kubectl and helm
    • Have a running Kubernetes cluster

Run locally in python virtual environment

Prerequisites: python3 and make installed.

  1. Get an envfile by running the following command.
mv .env.example .env
  1. Configure the .env file.

For Walrus backend,

  • Set TOOLKITS=walrus
  • Fill in OPENAI_API_KEY, WALRUS_URL and WALRUS_API_KEY

For Kubernetes backend,

  • Set TOOLKITS=kubernetes
  • Fill in OPENAI_API_KEY
  1. Run the following command to install. It will create a venv and install required dependencies.
make install
  1. Run the following command:
make run

Run with Docker

Prerequisites: docker installed.

  1. Get an envfile by running the following command.
mv .env.example .env
  1. Configure the .env file.

For Walrus backend,

  • Set TOOLKITS=walrus
  • Fill in OPENAI_API_KEY, WALRUS_URL and WALRUS_API_KEY

For Kubernetes backend,

  • Set TOOLKITS=kubernetes
  • Fill in OPENAI_API_KEY
  1. Run the following command:
docker run -it --env-file .env sealio/appilot:main

About

AI driven application management

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 97.9%
  • Makefile 2.0%
  • Dockerfile 0.1%