← All articles How it works

Running AI inside your environment: what self-hosting means in practice

Why the Insight products are built to deploy in your cloud accounts or on-prem, what stays inside your perimeter, what you still decide (models, identity, network), and questions to settle before a first deployment.

"Deployed inside your environment" appears on almost every Insight product's site. It's a deliberate design choice, and it's worth being precise about what it means, because it changes who does what in a first deployment.

What it means

The Insight products are built to run in your cloud accounts (AWS, Google Cloud or Azure) or on-prem, as containers alongside your data. The content they index, the traces and audit logs they write, and the metadata they collect stay in stores you own. People sign in through your identity provider, and access rules come from your roles and data classifications. Several product sites put it the same way: your data never leaves your perimeter.

What you still decide

  • Which models. Self-hosting the product doesn't decide where inference happens. You can point products at private model endpoints in your own cloud, at models you host, at domain models built with Insight DLM, or at external APIs. If you choose an external API, that's data leaving the perimeter, and it should be a conscious decision.
  • Identity and roles. Which groups can search which sources, approve which prompts, or see which traces.
  • Network and egress. Which connectors reach which systems, and what (if anything) may call out.
  • Retention. How long indexes, traces and audit logs are kept.

Why it matters

For regulated industries, the first questions about any AI system are where the data goes, who can see it, and whether you can prove afterwards what happened. Running the products inside your environment makes the first two answerable with your existing controls, and the audit and tracing built into the products answer the third.

Compliance is shared. Product sites describe their compliance-ready features, such as audit trails, encryption and access control. Whether a particular deployment meets HIPAA, GDPR, SOC 2 or another framework depends on how it's configured and operated in your environment, and is something to confirm with your own compliance team.

Questions to settle before a first deployment

  1. Where does it run? Which account, region and cluster; who operates it.
  2. Which sources first? Two or three, with an owner for each and a clear access model.
  3. Which models? And whether any data may leave the perimeter for inference.
  4. Who signs in, and how? SSO groups mapped to roles.
  5. What must be auditable? Answers, agent actions, prompt changes: decide before go-live, not after.
Example · a made-up health plan

A deployment checklist on one page

A made-up regional health plan wanted staff search over clinical policies and member-service runbooks. Before anything was installed, its platform team answered the five questions: one AWS account in its existing VPC, SharePoint and ServiceNow as the first two sources, a private model endpoint in the same account, SSO groups for member services and clinical review, and tracing on every answer kept for the plan's standard retention period.

Who does what

Self-hosting changes the shape of a deployment. A few responsibilities sit naturally with each side:

Your teamsThe product (and VerticalServe, if you ask us to stand it up)
Cloud account, network and clusterContainer images and deployment configuration
Identity provider, groups and rolesMapping roles to permissions inside the product
Source owners and access approvalsConnectors, ingestion and query-time enforcement
Model choice and any egress policySupport for private endpoints and self-hosted models
Retention and compliance decisionsAudit trails, tracing and the settings to enforce them

The product sites also describe the option of having VerticalServe stand the products up for you; either way, the infrastructure is yours.

Related: the reference architecture · governing prompts and agents.

See the right Insight products in your environment

Tell us what you're trying to build, and we'll walk you through the products that fit, with your data, where it lives.

Request a demo

Keep reading

Guide
Ask + RAG + LenseCatalog + PromptsDLM

Guide: a phased adoption plan for the Insight platform

Where to start and what to add next: a first phase with Ask, RAG and Lense, a second with Catalog and Prompts, and a third with domain models and domain solutions, with the questions that decide when to move on.

November 17, 2026 · 8 min read
Use case
Insight CatalogAutomation HubInsightKafkaInsight Dataflow

Use case: moving a data platform without losing track of what depends on what

A made-up retailer's data team migrates warehouses and Kafka clusters with Insight Catalog for lineage and impact analysis, Insight Automation Hub for validated migration runs, InsightKafka for cluster health, and Insight Dataflow for the new pipelines.

November 15, 2026 · 8 min read
Use case
Insight AskInsight RAGInsight Lense

Use case: one search box for IT operations, across tickets, wikis and code

How a made-up manufacturer's IT operations team uses Insight Ask on top of Insight RAG: connect ServiceNow, Confluence, Jira, Slack and Git, keep permissions intact, and trace every answer with Insight Lense.

November 13, 2026 · 7 min read