Integration

Configure AI model settings by tenant

iService stores model provider, embedding, and vector store settings for each tenant. Administrators review and replace provider settings. iService retrieves relevant interactions or articles and sends them to the configured model as grounding context. The model returns a synthesized response from that context.

An external AI model provider connected to a controlled service-context boundary

Choose a tenant AI model provider

AI architecture, security, and procurement teams select model providers, embeddings, and vector stores for each tenant.

iService lets administrators choose the model provider for procurement and data requirements. Each tenant stores provider settings, embeddings, and vector stores for AI-assisted features. Teams review and replace those settings while the tenant retains its AI-assisted features.

A dense ordered field of tiny points with one cluster glowing where a query has landed

Set the provider, embeddings, and vector store

Tenant settings store model provider access, embeddings, vector stores, keys, and embedding generation controls.

iService configures model access per tenant. The tenant controls provider settings, embeddings, and vector stores. Administrators replace a saved OpenAI key while iService keeps its current value hidden. Separate controls pause embedding generation while existing embeddings remain available.

iService configured against an OpenAI model through the external API settings

Ground a response in service records

An AI-assisted response uses relevant interactions or articles from the service record as grounding context.

An agent starts an AI-assisted feature for a customer question in iService. iService retrieves relevant interactions or articles and passes them to the configured AI model as grounding context. The model returns a synthesized response based on that context. Tenant provider, embedding, and vector store settings remain configured for the feature.

Provider settings separate from AI features

Tenant settings control model providers separately from the AI-assisted features that use grounding context.

iService stores model access in tenant settings. Administrators review and replace provider settings. The tenant-level setup keeps provider choice separate from the grounding context selected for a response and the AI-assisted features that use it.

Tenant controls for keys and embeddings

Administrators replace a provider key and pause embedding generation through tenant settings.

iService lets administrators replace a saved OpenAI key while keeping its current value hidden. Tenant controls pause embedding generation while existing embeddings remain available. The tenant retains its configured AI settings and stored embeddings.

Read AI model configuration docs

iService stores model provider, embedding, and vector store settings for each tenant and grounds responses in service records.

Read Model Configuration Docs