Multi-Tenant Data Platform to support all your clients

One place for all your clients' ELT pipelines and warehouses. Define your own templates, quickly onboard new clients, and override & extend the defaults to make it just right for them.

How it works

Client 1

Client 2

Client 3

SQL

API

Semantic Layer

Your Product

Made for productized services

Building and maintaining a dedicated data stack per client is slow and expensive when most of what they need is standard. Arch lets you take your common data solutions, deploy them to clients as a managed service, and easily customize them to their needs.

Step 1

Define your templates

Templates can be defined using versioned code or in the UI

Ingest from any source

Define custom tables using dbt Core models or Python, or easily apply row-level transformations like masking PII or dropping columns.

Transform as desired

Define custom tables using dbt Core models or Python, or easily apply row-level transformations like masking PII or dropping columns.

Store where you need it

Use Arch’s built-in multi-tenant data warehouse powered by Hydra, or export to an external data warehouse or lake.

Step 2

Onboard your clients

Clients can be onboarded manually, programmatically, or self-serve

Auto-provisioned

When you onboard a new client onto a specific template, Arch automatically sets up their ELT orchestration and a data warehouse.

White-label client portal

Clients can set up and manage their own connections using OAuth and map entities and properties to fit the template’s data model.

Fully automatable

To manage clients, connections, and mappings more efficiently, you can use the API or embed Arch into in your custom frontend.

Step 3

Customize as needed

Data integration and transformation can be customized for each client

Override models & metrics

Every templatized transformation step can be overridden to meet a client’s needs, and common overrides can be reused

Swap out sources

Support multiple competing SaaS apps (like CRMs) by defining a unified data model that downstream models can build on.

Add new sources & models

When a client needs something special, you can easily build it just for them and optionally add it to the template later.

Step 4

Access their data

Each client's data is readily available to your frontend of choice

SQL for BI & data science

Arch’s built-in data warehouse is powered by Postgres and Hydra, which are supported by every BI and Notebook tool.

API for custom frontends

Every table in Arch can be exposed as a REST or GraphQL API to power your custom web or LLM chatbot frontend.

Webhooks for external actions

Whenever a table gets new data, Arch can forward it to an external system using a webhook or API request (reverse ETL).

The latest from the Arch blog

Arch updates, data consulting trends, tutorials, and other data musings