THE meloki PLATFORM

One AI platform. You stay in control.

Combine local AI models, included standard solutions and your own applications on an expandable appliance – centrally managed with the AdminPACK.

CONNECTIONS, NOT ISOLATED SOLUTIONS

One platform. Your combination.

On one appliance. Centrally managed. Open to your solutions.

Your teamUse AI in everyday work
Data & systemsConnect your IT
Your meloki platform
IncludedYour choice

Standard solutions

  • Open WebUI
  • Ollama
  • n8n
  • Whisper
More standard solutions

Custom & additional
Solutions

Add flexibly
meloki AdminPACKInstall · Manage · Extend
Your local appliance
Standard solutions

Included. Your choice.

Use only what suits you. Add more services and your own applications as needed.

Ollama

Run local language models

OpenWebUI

Chat and work with company knowledge

n8n

Set up workflows and interfaces

Whisper STT

Convert speech into text locally

Docling

Prepare documents for further processing

Qdrant

Store vectors for semantic search

Kokoro TTS

Set up speech output as an additional service

Add your own solution

Room for your requirements.

Explore integration

Additional standard solutions are also included.

LOCAL MODELS

Let the task guide your model choice.

With local inference, an AI model generates its answers on your hardware. Choose the model that fits your tasks and resources.

Ollama provides local language models for tasks such as text, code or knowledge questions. You can change models later and add more.

Understand model selection →

A selection of our standard models

  • Gemma 4 (Gemini)
  • Devstral (Mistral Vibe)
  • GPT-OSS (ChatGPT)

Other models are available alongside these examples.

Choose the model and variant to suit your task, licence and hardware.

YOUR KNOWLEDGE AS THE FOUNDATION

The answer starts with your sources.

The AI searches your approved documents for relevant content and uses it to answer. This approach is called RAG.

  1. Your documents
  2. Find knowledge
  3. Answer

Make knowledge available

Documents feed into a searchable index. When sources change, the knowledge base needs to be maintained accordingly. Quality, permissions and freshness remain essential.

Assess and review

Uploading documents does not automatically train the language model. RAG adds context. Answers can still contain errors and should be checked against the sources.

INTEGRATION & AUTOMATION

From an answer to the next step.

Use n8n, APIs and webhooks to connect approved systems and local AI in a workflow.

An event triggers a workflow. AI can help classify information or draft text. Human approval can be included before a write action. Access, permissions, error handling and data flows need to be configured appropriately.

Plan your integrations →
  1. Enquiry
  2. AI draft
  3. Review & approve
  4. Create ticket
SPEECH & AUDIO

Spoken knowledge. Ready to use as text.

Whisper STT converts provided audio files into text locally. You can then summarise the text or extract possible tasks from it.

Recording permissions, consent and suitable file formats are part of the preparation. A direct connection to your conferencing system requires its own integration. Kokoro TTS can be configured as an additional service for speech output.

See the use case →
meloki AdminPACK

One central interface for your entire AI environment.

The AdminPACK manages the technical foundation of your appliance. Access and permissions within individual applications are configured in each application.

meloki AdminPACK

Your solutions.

Install, configure and start – in one place.

Your own solution

Open WebUI

Running

Chat and use approved documents.

Access via subdomainchat.meloki.local
Service details

Install and configure container services, and group related components into solution bundles.

  • Add, activate and deactivate solutions
  • Keep track of versions and dependencies
  • Configure routing via subdomain, path or port

You approve the update.

Recommended versions or your own choice.

Open WebUIRecommended version
Update available
InstalledRecommended
OllamaRecommended version
Up to date
n8nCustom version

Update with control.

  1. Choose version
  2. Back up first
  3. Install update
Update details

Use recommended releases or choose custom container versions. You control the update process.

  • Check and install available versions
  • Set up automatic checks or installations
  • Plan backups before updates

Everything in view.

See resources, service status and logs together.

CPU18 %
RAM38 %
GPU42 %
Storage61 %

GPU usage

Last 5 minutes
−5 min.Now

Service logs

OllamaModel ready

Open WebUIConnection established

Status and events
Monitoring details

Keep track of CPU, RAM, storage and GPU usage, and explore events and service logs.

  • Monitor GPU memory and temperature
  • View resource usage history
  • Search events and solution logs

Keep a working state.

Back up solutions and manage restore points.

Restore points

  1. Today · 08:00Chat & RAG
    Compatible
  2. Yesterday · 08:00Chat & RAG
    Compatible
Included solutions
Open WebUI+Ollama+ Nginx

Backup destination

LocalOn your appliance

More destinations after setup:
SFTP or S3

Backup details

Create restore points for your solutions and configure suitable backup destinations.

  • Local backups and recovery
  • Scheduled backups
  • SFTP or S3 destination after setup

Your access. Protected.

Password, one-time code and HTTPS for administration.

AdminPACK access

Your administration.
Your access protection.

PasswordFor the meloki user
Set
Two-factor protectionTime-based one-time code
HTTPS certificateYour own TLS certificate
Added

Configure permissions for Open WebUI and other services within each application.

Security details

Protect AdminPACK access with a password, an optional time-based one-time code and your own TLS certificates.

  • Set up two-factor authentication for the meloki user
  • Add your own HTTPS certificates
  • Configure access permissions within each application

Manage services centrally

Install and configure container services, and group related components into solution bundles.

  • Add, activate and deactivate solutions
  • Keep track of versions and dependencies
  • Configure routing via subdomain, path or port

Updates on your terms

Use recommended releases or choose custom container versions. You control the update process.

  • Check and install available versions
  • Set up automatic checks or installations
  • Plan backups before updates

See resources and status

Keep track of CPU, RAM, storage and GPU usage, and explore events and service logs.

  • Monitor GPU memory and temperature
  • View resource usage history
  • Search events and solution logs

Back up and restore

Create restore points for your solutions and configure suitable backup destinations.

  • Local backups and recovery
  • Scheduled backups
  • SFTP or S3 destination after setup

Secure administrative access

Protect AdminPACK access with a password, an optional time-based one-time code and your own TLS certificates.

  • Set up two-factor authentication for the meloki user
  • Add your own HTTPS certificates
  • Configure access permissions within each application
SECURITY WITH CLEAR RESPONSIBILITIES

Local means closer to your rules.

Your own appliance supports control over processing, access and operations. No public AI cloud is required for core local AI processing.

You decide which sources are connected and who can use them. Network rules, updates, backups and a permissions plan remain part of your operations.

Design data flows deliberately

  • Local models process content on your appliance
  • Your own TLS certificates and an optional one-time code secure administrative access
  • External APIs or S3/SFTP backup targets can transfer data
  • Privacy and legal requirements must be assessed for your specific use
Get more answers →
YOUR NEXT STEP

Explore meloki in a personal demo

Discover in a personal demo how meloki can support your company knowledge, applications and workflows.

Arrange a demo
The meloki mascot flies towards you