Journey-native analysis
Funnels, paths, attribution: sequence is the native unit.
Did A happen before B, how long between them, where did people go instead: those answers come back at conversational speed, because Alethiom’s engine stores journeys in sequence order. Change a step, re-order, add a breakdown: the answer updates while the question is still in your head.
- Funnels with step-level breakdowns and time-to-convert
- Path exploration forward and backward from any event
- Attribution models on any metric, over the whole journey, not the last click
drop file at site/media/prod-funnel.mp4
The semantic layer
Defined in SQL. Applied at query time. Retroactive to your first event.
A new metric, or a fix to an old one, is live across all of history the moment you save it, because metrics and dimensions are SQL definitions evaluated at query time. No pipeline change, no backfill. Or skip the SQL: describe what you mean, and the AI authors the definition.
- Real SQL, not a restricted formula language
- Written against your warehouse schema as it is, with no mapping into a vendor data model
- Journey-aware: a value can persist across a session or a person’s history
- One governed definition, used by every teammate, human and AI
drop file at site/media/prod-semantic-layer.mp4
Identity mapping
One journey per person, across every system that saw them.
Each person becomes one sequenced journey across web events, product events, transactions, and CRM touches, even though your warehouse holds them under different keys. Bring your own identity mapping table and Alethiom maps everything to your golden customer ID, with the rules visible and editable.
- Cross-device and cross-system resolution on your keys
- Identity-map updates restate history on the next cycle, with no lookback limitations
- Re-map at any time; the serving layer rebuilds
Dashboards & boards
A canvas where analyses live together and stay live.
Change a date or a segment once and every analysis on the board follows: funnels, trends, tables, and charts sit as blocks on an infinite canvas, wired to shared frames for date and segment scope. A dashboard distills a board down to what your executives read, and sits pinned on the home page.
- Blocks share scope through frames: one change, and every block follows
- From exploration to executive dashboard without rebuilding
- Every number traceable back to its query
shares site/media/home-board.mp4 with the home page
The AI teammate
The AI can do anything you can do. Nothing you can’t.
Ask in plain language and the AI answers with real artifacts (funnels, boards, analyses) through the same governed layer as the interface. It can do anything you’re allowed to do: run analysis, check data quality, even file product feedback. Every answer is a traceable query you can open, inspect, and re-run.
- Summon it anywhere in the product, and it answers in the context you’re looking at
- Same permissions, same governance as the person asking
- Every figure links to the query that produced it
- Your own agents can run the same reports over MCP, under each person’s permissions
- One switch turns it off, and you can use your own Anthropic API key: AI on your terms
drop file at site/media/prod-ai.mp4
The briefing
The AI works the night shift.
Every night the AI investigates your data (what moved, what broke, what’s quietly trending), and a briefing is waiting on your home page by morning. Every finding shows where it came from, and every number links to the query behind it. Tell it what you care about and the next briefing listens.
- Investigated overnight, waiting before you log in
- Personal to each teammate: your projects, your dashboards, your context
- Findings come with receipts: every number re-checked and traceable
- Open any finding into a conversation and keep digging
