Pay-per-answer

Create natural language answers to user questions based on your raw Corporate KnowledgeBase. This plan uses our curated LLM and does not offer connectivity to other LLMs. All other features are available with this plan. The BYOLLM plan is available if this is a desired feature.

This plan's features include, but are not limited to:

• Automatic catalog, curation and grouping of questions and answers
• Export to a Virtual Agent
• Round-trip monitoring to a Virtual Agent
• Sentiment scoring
• Automatic language detection
• Translate text into other languages
• Extract Entities from text
• Categorize text by matching categories and matching or creating Intents

Flex

The NeuralSeek Flex plan is a bring-your-own LLM plan featuring unlimited usage, and a flex license allowing you to optionally and additionally install NeuralSeek components on your hardware, behind your firewall as needed to meet your security requirements while you are subscribed to this flex plan. All NeuralSeek features are supported on this plan.

Each base instance (or install) is licensed for 10,000 users. Additional users may be added in blocks of 10,000.

On-Premise Details

The Flex plan allows you to optionally and additionally install NeuralSeek on your hardware, behind your firewall to meet security requirements.

Installation Requirements

Minimum sizing requirements for on-prem installation include:

• 12 Core CPU
• 64 GB RAM/Memory
• 100 GB Available Disc Space
• If self-hosting an LLM (not using watsonx.ai or sagemaker) your self-hosted LLM will require a GPU VM that is equivalent or better to a single NVIDIA A10G

Installation Steps

1. Log onto Red Hat OpenShift console with appropriate domain.
2. Modify the appropriate .yml file with the corresponding hostname OpenShift external URL.
   • .yml files are provided during consultation meeting.
3. Verify connectivity to the Cerebral Blue docker in .yml files.
   • Permission access will be granted during consultation meeting. Provide the appropriate username.
4. Copy the contents of the .yml files into your OpenShift console by clicking the plus icon, then click create.
5. Route will be created manually by navigating to Networking → Routes → Create Route.
   • Add a unique name.
   • Select the service to route to.
   • Select the target port for traffic.
   • Optionally, provide a TLS certificate. Default will set to HTTP.
6. Click the link to the route to open the NeuralSeek User Interface.

Bring-your-own-LLM

Leverage all of NeuralSeek's features, but instead of using our curated LLM, you can connect via our no-code connectors to leading commercial and open-source LLM's. This enables you to run within a single datacenter or country, or choose the commercial LLM that best fits your business and pricing needs.

Refer to our Integrations documentation for a list of supported LLM's.

Search

The Search Plan is for use cases not requiring a Virtual Agent. NeuralSeek provides a search interface to supported KnowledgeBases, and will provide search responses plus generative AI summaries. Any generated AI summary incurs a per-call usage fee. Cache responses are included at no additional cost.

This plan's features are identical to the pay-per-answer plans EXCEPT:

• No export to a Virtual Agent is allowed
• No round-trip monitoring to a Virtual Agent is allowed
• No sentiment scoring
• No automatic language detection

Small Business

The Small Business plan is the easiest plan to get NeuralSeek running in minutes with no experience required. This plan does not require connection to an external KnowledgeBase. Simply point NeuralSeek at your website or upload documents, connect to a Virtual Agent, and go-live!

 

 

For cloud-specific available plans, see cloud provider for up-to-date cost information.