Intelligence Configuration Guide (for Streamlined Search and AI Enrichment)

The Intelligence section in the Control Center lets you integrate large language models into your workflows. You can choose between using models hosted through Ollama (for local and secure inference ) or OpenAI (for cloud-hosted models via API). This allows you to enrich data, extract structured insights, and power AI-based features across the platform (i.e. Streamlined Search Document Q&A).

Choosing a Model Provider

At the top of the Intelligence settings panel, you'll find the Choose model dropdown. You can select from:

Ollama: For using locally hosted models.
OpenAI: For using models from the OpenAI API.

Ollama Model Setup

To use an Ollama-hosted model, select Ollama in the dropdown.

Then fill out the fields:

Name: A friendly name to identify this model (e.g. Desktop-llama3:8b)
Server URL: URL to your Ollama instance (typically http://localhost:11434)
Text Model: The name of the language model to use (e.g. llama3:8b, gemma:2b)
Embeddings Model: Model name used for generating embeddings (e.g. nomic-embed-text).
- This option is not optional for models configured to do document Q&A. Embeddings are generated at that time, and a model must be selected.
- Shinydocs highly recommends using nomic-embed-text:latest as your embeddings model for performance.

Optional Advanced Fields

These settings give you fine-grained control over how the model behaves. They’re not required, but adjusting them can help you balance creativity, speed, and consistency depending on your use case. If you're not sure what to change, the defaults are a good starting point.

Option	Description	Usage
Seed	Random number used to get repeatable responses from the model.	Use if you want the same result every time for a given prompt.
Temperature	Controls randomness. 0 = more predictable, 1 = more creative.	Use 0.2–0.5 for reliable responses, 0.7+ for brainstorming.
Max Supported Tokens	The max tokens for input + output. Defaults to 4096 for text, 8192 for embeddings.	Match this to your model's actual context limit.
Mirostat Sampling	Algorithm for adaptive sampling. Can be set to `Disabled`, `Mirostat`, or `Mirostat v2`.	Only change if you're experimenting.
Mirostat ETA	Learning rate for Mirostat feedback. Default is 0.1.	Lower = slower learning, higher = faster.
Mirostat Tau	Coherence vs. diversity. Lower = more focused output.	Try 5.0 to start.
Context Window Size	The total tokens the model can "see" at once.	Match your model's capability. E.g. `32768` for llama3:8b.
Repeat Last N	Prevents repetition. 0 = off, 64 = standard.	Helps avoid loops in responses.
Tail Free Sampling	Reduces the impact of less probable tokens. Default is 1.0.	Lower it for more conservative output.
Tokens To Predict	Max tokens to generate. `128` is a good default. Use `-1` for infinite generation.	Adjust if you need longer or shorter responses.
Top K	Limits token options to top K most likely. Higher = more diversity.	40 is a balanced default.
Min P	Filters low-probability tokens. 0.0 = no filtering.	Increase to make the output more precise.
Max Batch Size	Requests handled in parallel. 1 = safest.	Increase only if you know your setup can handle it.

OpenAI Model Setup

At this time, AI Enrichment (content tagging) is not supported with OpenAI models.

If you're using the OpenAI API, choose OpenAI in the dropdown.

Required Fields

Name: Name for the integration (e.g. GPT-4 Cloud)
Server URL: Leave this blank unless you're using a proxy.
API Key: Your OpenAI API key (starts with sk-)
Text Model: The name of the model (e.g. gpt-4, gpt-3.5-turbo)
Embeddings Model (optional): The model used to generate embeddings (e.g. text-embedding-ada-002)

Once saved, the OpenAI model will appear in the Model List.

Model List and Managing Defaults

At the bottom of the page, you’ll see a Model List table with all configured models. You can:

Set a model as default: This model will be used unless another is explicitly selected.
Check which model is active: Only one model runs at a time, and is marked as Active.
Add multiple models: Great for switching between testing and production setups.

Using, Switching, or Updating Models

In the Intelligence configuration, you have a two options available:

Set default
Edit

Set default

The intelligence engine set as default will be the configuration used in Shinydocs Streamlined Search AI. When a user asks questions about documents with the AI chat bot, this is the configuration that will be used.

To switch models, click Set default on another model from the list. If you update the Ollama backend (e.g. swap to a newer version of a model), you may want to remove and re-add the model entry here to match.

AI Enrichment (content tagging) has the option to select which intelligence engine to use per prompt, and is not locked to the default engine.

Edit

You can edit existing intelligence engines to tweak advanced model options, the Ollama URL, and more by clicking on the kebab menu (3 vertical dots) for the respective intelligence engine.

UI showing configured intelligence engines.

You can also delete an existing engine by clicking Delete.