Questions and answers

Getting started

What tags do I need to implement?

None. Alethiom reads the tables already in your warehouse, in the schema they already have. If your web and app events aren’t in your warehouse yet, or you want to move off your analytics vendor’s collection and collect straight to the warehouse, any method of getting them there works.

Snowflake, BigQuery, and Databricks each have a streaming write API that a server-side tag or a small endpoint can post your data layer to: BigQuery’s Storage Write API, Snowflake’s Snowpipe Streaming, Databricks’ Zerobus Ingest.

There are also products that run the collection pipeline for you, such as Hightouch Events, RudderStack, or Snowplow.

What if I need to combine the historical data from my analytics tool with the new data I collect?

Alethiom can ingest the historical data and the new data into a single dataset, giving you one continuous view. You can map the historical schema onto the new one in your warehouse before Alethiom pulls it, or in Alethiom when you bring the table in. Alethiom recognizes GA4 BigQuery exports and Adobe Analytics data feeds and sets them up automatically.

Which warehouses are supported?

Snowflake, BigQuery, and Databricks today. Support for other warehouses can be added on request.

What kinds of data can it use?

Three kinds of tables today. Event tables: one row per event (e.g. web and app events, orders, loyalty activity, support tickets, call-center logs), with a timestamp and a column that identifies the person (a device, anonymous, or customer ID).

Lookup tables: reference data keyed on a column your events carry (a product catalog, a store list, campaign names). Person attributes fall into this category, as a lookup keyed on the person ID.

An identity-mapping table: the anonymous IDs in your event data alongside the known customer ID (or IDs) you want them restated as.

What does it cost?

You pay your own infrastructure costs, at the rates you’ve negotiated with your cloud provider, plus a flat yearly product fee sized to the implementation we manage for you. There is no per-query charge: exploration runs on Alethiom’s engine in your account, not on your warehouse.

Does it replace our digital analytics tool?

Yes. Alethiom is designed to replace traditional vendor SaaS analytics products for the journey workload: funnels, paths, attribution, and segments on the complete record in your warehouse.

How do we try it?

A 30-minute walkthrough, a guided pilot on sample data, or both. You try everything before anything touches your account.

Data and security

Does any of my data leave my cloud account?

No. The whole product runs inside your account, and query results go from your cloud straight to your users’ browsers.

Where is the data stored?

In your cloud account, in the region you choose. The warehouse stays the system of record, and Alethiom’s cache sits beside it in the same account.

Do reports in Alethiom run queries in my warehouse under the hood?

No. Only a small scheduled sync query touches your warehouse, as often as every five minutes. Interactive queries run against a disposable cache in your account, rebuildable from the warehouse at any time.

Who operates it?

We do, through a scoped role you grant and can revoke at any time. You host the stack, but you don’t have to operate, upgrade, or maintain it.

Can the app be private, with no public internet exposure?

Yes. You choose at deployment: internal-only, reached through your VPN or your cloud’s private connectivity, or public behind TLS.

Is our data isolated from other customers?

Yes. The whole stack runs inside your account, so there is no shared database or shared service holding several customers’ data.

Is the data encrypted?

TLS in transit, and at rest under your cloud’s storage encryption. Warehouse credentials are stored encrypted inside the cluster in your account and never leave it.

How do we sign in?

Through your identity provider. Sign-in federates to your SSO, with your MFA policy and your deprovisioning, and there is no separate password store.

What happens if we end the contract?

Revoke the role and we are locked out. Your warehouse is untouched, and Alethiom’s cache in your account is disposable, so there is nothing of yours to migrate out.

AI

Does the AI train on our data?

No. Model calls run under Anthropic’s commercial API terms, which exclude training. The AI service runs in your account, one switch turns it off, and off means zero model-provider traffic. You can also bring your own Anthropic key.

Can our own AI tools and agents use it?

Yes, over MCP. Users sign into an MCP client as themselves and run reports under their own permissions, on the same definitions and engine the interface uses.

Product

How does it connect one person across systems?

With your identity-mapping table. Each person becomes one sequenced journey across web events, product events, transactions, and CRM touches, mapped to your customer ID, with the rules visible and editable. Change the mapping and history restates on the next data sync.

Will the numbers match our warehouse?

The raw data matches your warehouse by construction. The dimensions, metrics, and segments you define in Alethiom are another matter: they may not be easy to reproduce in warehouse SQL.