mAItion

mAItion is an all-in-one ready-to-use AI-powered tool that combines your existing knowledge with LLMs, allowing you to chat, search and interact with your data through a slick chat interface. With mAItion you can aggregate all your knowledge from many sources using Connectors into a central place and interact with your knowledge with ease!

✨ Features

• Support for both local and remote LLMs for embedding and inference
• Asynchronous data ingestion with deduplication and per-source configurable schedules
• Data ingestion from S3 buckets with Everything-to-Markdown conversion via MarkItDown
• Data ingestion from MediaWiki with Wiki-to-Markdown conversion via html2text
• SerpAPI search query results ingestion from Google Search results with customizable queries
• Flexible configuration supporting an arbitrary number of connectors
• Built with extensibility in mind, allowing for custom connectors addition with ease
• MCP servers support (stdio, streamable http)
• Web-search support (through external services and via on-premise services)
• In-place chat with uploaded documents and images (for multi-modal LLMs)
• Code execution and Code interpreter
• Text-to-Speech and Speech-to-Text capabilities
• Image generator (requires model supporting image generation)
• Flexible automation capabilities through Functions and Pipelines
• Multi-user setup with fine-grained groups and permissions
• Support for multiple customized configurations of LLM models

✨ Use-cases

• A single place to chat with your company knowledge that's scattered across many external systems
• A central space for looking up and refining your existing knowledge across many knowledge bases
• A tool to find secret knowledge that can not be found in the other was across your scattered data
• An entry-point into your on-premise hosted LLM models supporting evaluations and per-model settings

🌐 Connectors included

• S3 (any AWS compatible Object Storage including AWS, Contabo, B2, Cloudflare R2, OVH, etc)
• MediaWiki (all versions supported, both private and public wiki)
• SerpAPI

🌐 Extra connectors

Over 100 extra connectors are available at request, including the most popular ones:

• Gmail
• Google Drive
• Jira
• Slack
• GitHub
• Gitlab
• Notion
• Microsoft Teams
• Microsoft Office 365
• Dropbox
• Trello
• YouTube
• FTP
• SCP
• SSH
• and many more..

Quick start

Requirements

• Docker and Docker Compose
• OpenRouter or OpenAI API key (a free OpenRouter account works with the default configuration)
• S3 bucket

Setup

• Create .env.rag file by copying .env.rag.example (see https://github.com/wikiteq/rag-of-all-trades for details)
  • Set OPENROUTER_API_KEY
  • Set S3_ACCOUNT1_* values to match your source S3 bucket with files
• Create config.yaml out of config.yaml.example
  • The default config works OK and is configured to:
    • Use a single S3 bucket as data source
    • Use openai/gpt-oss-20b:free model for rerphrase
    • User local sentence-transformers/all-mpnet-base-v2 model for embeddings
    • You can change the values if necessary, refer to https://github.com/wikiteq/rag-of-all-trades for details
• Create .env file by copying .env.openwebui.example
  • Set OPENAI_API_KEY
  • Optionally set OPENAI_DEFAULT_MODEL

Start the stack by running docker compose up -d. Wait until all the services become healthy. You can check health status by running docker compose ps and checking the STATUS column of the services:

docker compose ps
NAME                                      IMAGE                                      COMMAND                  SERVICE     CREATED         STATUS                   PORTS
rag-of-all-trades-openwebui-api-1         ghcr.io/wikiteq/rag-of-all-trades:latest   "sh -c 'alembic upgr…"   api         4 minutes ago   Up 4 minutes (healthy)   8000/tcp
rag-of-all-trades-openwebui-openwebui-1   ghcr.io/open-webui/open-webui:0.6.5        "/custom-entrypoint.…"   openwebui   4 minutes ago   Up 4 minutes (healthy)   0.0.0.0:3000->8080/tcp, [::]:3000->8080/tcp
rag-of-all-trades-openwebui-postgres-1    ankane/pgvector:v0.5.1                     "docker-entrypoint.s…"   postgres    4 minutes ago   Up 4 minutes (healthy)   5432/tcp
rag-of-all-trades-openwebui-redis-1       redis:7                                    "docker-entrypoint.s…"   redis       4 minutes ago   Up 4 minutes (healthy)   6379/tcp

It takes up to a minute for the OpenWebUI to fully boot on cold start.

Once all the services are booted and report healthy status visit http://localhost:3000 and login using Admin credentials. The credentials are defined in X_WEBUI_ADMIN_EMAIL and X_WEBUI_ADMIN_PASS of the .env file. The default ones are:

• username: admin@example123.com
• password: q1w2e3r4!

If you did not change the ENABLE_OPENAI_API you will also have LLM provider pre-configured with the values you have in the .env including the default chat model

The filter function that's responsible for the RAG service communication will also be automatically provisioned and enabled globally. You can change these settings at the Admin panel

Connectors configuration

The service supports multiple data sources, including multiple data sources of the same type, each with its own ingestion schedule. The connectors to enable are defined via config.yaml, and their secrets are defined in the .env.rag file.

S3 Connector

The S3 connector ingests documents from S3 buckets and converts them to Markdown format. The connector has the following configuration options:

# config.yaml

sources:
  - 
  - type: "s3" # must be s3
    name: "account1" # arbitrary name for the connector, will be stored in metadata
    config:
      endpoint: "${S3_ACCOUNT1_ENDPOINT}" # s3 endpoint
      access_key: "${S3_ACCOUNT1_ACCESS_KEY}" # s3 access key
      secret_key: "${S3_ACCOUNT1_SECRET_KEY}" # s3 secret key
      region: "${S3_ACCOUNT1_REGION}" # s3 region
      use_ssl: "${S3_ACCOUNT1_USE_SSL}" # use ssl for s3 connection, can be True or False
      buckets: "${S3_ACCOUNT1_BUCKETS}" # single entry or comma-separated list i.e. bucket1,bucket2
      schedules: "${S3_ACCOUNT1_SCHEDULES}" # single entry or comma-separated list i.e. 3600,60
      
  - type: "s3"
    name: "account2"
    config:
      ...

  - type: "s3"
    name: "account3"
    config:
      ...

# .env.rag

S3_ACCOUNT1_ENDPOINT=https://s3.amazonaws.com
S3_ACCOUNT1_ACCESS_KEY=xxx
S3_ACCOUNT1_SECRET_KEY=xxx
S3_ACCOUNT1_REGION=us-east-1
S3_ACCOUNT1_USE_SSL=True
S3_ACCOUNT1_BUCKETS=bucket1,bucket2
S3_ACCOUNT1_SCHEDULES=3600,60

MediaWiki Connector

The MediaWiki connector ingests documents from MediaWiki sites and converts them to Markdown format. The connector has the following configuration options:

# config.yaml

sources:
  - type: "mediawiki"
    name: "wiki1"
    config:
      api_url: "${MEDIAWIKI1_API_URL}"
      request_delay: 0.1
      schedules: "${MEDIAWIKI1_SCHEDULES}"

  - type: "mediawiki"
    name: "wiki2"
    config:
      ...

  - type: "mediawiki"
    name: "wiki3"
    config:
      ...

# .env.rag

MEDIAWIKI1_API_URL=https://en.wikipedia.org/w/api.php
MEDIAWIKI1_SCHEDULES=3600

SerpAPI Connector

The SerpAPI connector ingests documents from Google Search results and converts them to Markdown format. The connector has the following configuration options:

# config.yaml

sources:
  - type: "serpapi"
    name: "serp_ingestion1"
    config:
      api_key: "${SERPAPI1_KEY}"
      queries: "${SERPAPI1_QUERIES}"
      schedules: "${SERPAPI1_SCHEDULES}"

  - type: "serpapi"
    name: "serp_ingestion2"
    config:

  - type: "serpapi"
    name: "serp_ingestion3"
    config:

# .env.rag

SERPAPI1_KEY=xxxx
SERPAPI1_QUERIES=aaa
SERPAPI1_SCHEDULES=3600

Embeddings and Inference

Embeddings support

Both local and remote OpenAI-compatible models are supported for embeddings:

• Local (running arbitrary embedding models from HuggingFace)
• OpenRouter
• OpenAI or OpenAI-compatible

Inference support

Both local and remote OpenAI-compatible models are supported for inference:

• OpenRouter
• OpenAI or OpenAI-compatible

Embeddings-only HuggingFace local model

You can configure the service to use local embeddings only, in this mode you can use any embedding model supported by HuggingFace. Inference is disabled in this mode, so you won't be able to use the rephrase endpoint.

# config.yaml

embedding:
  provider: local
  # you can use any embedding model supported by HuggingFace
  model_config: sentence-transformers/all-MiniLM-L6-v2
  embedding_dim: 384

inference:
  provider: None
  model_config: None

Embeddings-only OpenRouter/OpenAI model

You can configure the service to use remote embeddings, in this mode you can use any embedding model supported by OpenRouter/OpenAI. Inference is disabled in this mode, so you won't be able to use the rephrase endpoint.

# config.yaml

embedding:
  provider: openrouter
  model_config: text-embedding-3-small
  embedding_dim: 1536

inference:
  provider: None
  model_config: None

You must set OPENROUTER_API_KEY and OPENROUTER_API_BASE in the .env.rag file.

Embeddings and inference OpenRouter/OpenAI model

You can configure the service to use remote embeddings and remote inference, in this mode you can use any embedding and inference models supported by OpenRouter/OpenAI.

# config.yaml

embedding:
  provider: openrouter
  model_config: text-embedding-3-small
  embedding_dim: 1536

inference:
  provider: openrouter
  model_config: gpt-4o

You must set OPENROUTER_API_KEY and OPENROUTER_API_BASE in the .env.rag file.

Reference of the config.yaml

The config.yaml file contains the main configuration of the service.

Environment variables (${...}) in the config file are evaluated at runtime.

sources: # holds the list of sources to ingest from (Connectors)

  - type: # type of the connector (s3, mediawiki, serpapi)
    name: # arbitrary name for the connector, will be stored in metadata
    config:
      # connector specific configuration
      schedules: "${S3_ACCOUNT1_SCHEDULES}"

# configures models and dimensions for embeddings
embedding:
  provider: openrouter # `openrouter`/`openai` or `local` for local HuggingFace embeddings
  model_config: text-embedding-3-small # model to use
  embedding_dim: 1536 # dimensions (check with the model docs)

# configures the LLM provider and model
inference:
  provider: openrouter # `openrouter`/`openai`
  model_config: gpt-4o # model to use

# vector store configuration
vector_store:
  table_name: embeddings
  hybrid_search: true # whether to use hybrid search or not
  chunk_size: 512 # chunk size for vector indexing
  chunk_overlap: 50 # overlap between chunks
  # hnsw indexes settings
  hnsw:
    hnsw_m: 16 # number of neighbors
    hnsw_ef_construction: 64 # ef construction parameter for HNSW
    hnsw_ef_search: 40 # ef search parameter for HNSW
    hnsw_dist_method: vector_cosine_ops # distance metric for HNSW

Tech Stack

• Rag-Of-All-Trades v0.1 as a RAG backend
• OpenWebUI v0.6.5 as a front-end

Troubleshooting

OpenWebUI does not start

The openwebui service depends on the api service healthiness and will remain pending until the API service is online. Check the api container for any errors, review the config.yaml and .env.rag for typos.

HuggingFace connection timeout

requests.exceptions.ReadTimeout: (ReadTimeoutError("HTTPSConnectionPool(host='huggingface.co', port=443): Read timed out. (read timeout=10)"), '(Request ID: da122313-e11f-4d54-b4f3-187abfea0ca3)')

OpenWebUI downloads some HuggingFace models during first boot. Sometimes HuggingFace endpoints may timeout. In this scenario just run docker compose down -v to wipe the stack and start over with docker compose up -d