§ Promoted · noindex Analytics is now a platform-layer feature — every module is analytics-native, not a separate module.
Page preserved for reference · 2026-05-10
Home· Govern· Analytics & Insights
Analytics & Insights · M.04

The number. Drilled to the truck roll.

Cross-loop analytics that read from the same source as every Allometry module. Drill from a P&L line to a job to a crew to an asset to a contract — without a single warehouse-side join.

cube.live · revenue + ops + risk 28 DIMENSIONS · 4 LOOPS · ZERO ETL
Q3 cube · drill from any line
$84.2M · revenue · 22.4% margin · 91% NRR
S-04 segment · margin 28%
drill: 64 jobs · 12 crews
Atlanta region · margin 14%
drill: vendor B + crew C
Renewals · NRR 91% · cohort A
drill: contract terms
Fleet B uptime 99.6%
drill: per-asset MTBF
What it looks like

KPIs. Trends. Lineage.

Board-grade KPI tiles. Annotated trends with the moves that mattered. Drill from board number to truck roll.

§ 01 · KPI tiles

14 KPIs · all loops

Q3 2026 · 6 of 14
BOOKINGS
$24.6M
MARGIN
32.4%
NRR
128%
CAC PAY
11mo
§ 02 · Annotated trend

Margin · 12 quarters

Auto-tagged from event log
Q1: Margin Protect rollout · Q3: Cost Engine v2
§ 03 · Drill lineage

Board → truck roll

Q3 margin · 32.4%
L0 Board · 32.4%
L1 Segment · ENT 34.2%
L2 Region · East 35.1%
L3 Deal · QT-0892
L4 Crew · B-04close
The problem

Your dashboards can't drill to the truth.

Revenue lives in Tableau. Ops lives in Looker. Asset performance lives in a vendor portal. The CFO asks 'why did Atlanta's margin collapse' and four teams pull four different reports that don't reconcile.

Without Analytics & Insights

Today's status quo

  • Each function has its own BI tool, its own warehouse extracts, its own joins
  • Cross-loop questions ('why is Atlanta margin down') take 2 weeks of analyst time
  • Drill stops at the dashboard — to go deeper, dump CSVs, open Excel
  • QBR slides hand-built; numbers reconcile by Tuesday, presentation Thursday
With Analytics & Insights

What changes

  • One cube spans revenue, ops, asset, risk — drill freely across loops
  • Cross-loop questions answer in seconds — natural language or click-through
  • Drill from any aggregate to the source job, truck roll, contract clause, asset
  • QBRs assemble themselves — narrative + chart + drill embedded
Capabilities

What's inside.

The six capabilities that make this module work end-to-end. Pick any one as your starting point — they compound.

01

Cross-loop cube

28 dimensions across 4 loops in a single semantic model. Customer × asset × job × crew × region × period × scenario × policy. Drill anywhere.

Cube · Semantic
02

Source-truth drill

From any aggregate — drill to source. From a P&L line to a job to a truck roll to a part to a contract clause. The path is always there.

Drill · Source
03

Natural-language Q&A

Ask in plain English. The cube knows the schema, the metrics, the relationships. Answers come with the drill paths attached.

NLQ · Q&A
04

Scenario modeling

What-if a 6% wage increase? A vendor change? A pricing test? Modeled across the cube — revenue, ops, risk all updated together.

Scenario · What-if
05

Auto-narrative

Quarterly reviews assemble themselves — the numbers, the why, the drill, the recommended action. Edited by humans; not built from scratch.

Narrative · QBR
06

Embedded BI

Embed reports inside Allometry workflows — the AE sees their own pipeline analytics inside Quoting Engine. Insight where work happens.

Embed · In-flow
The autonomous loop

From question to drill path.

Question lands on the cube, semantic model resolves, answer renders, drill path attaches. From aggregate to truck roll in two clicks. Without ETL.

§ 01 · Ask

Question or click

Natural language, or click-through from any module's chart. Same cube, same answer.

< 1s
§ 02 · Resolve

Semantic match

Metrics, dimensions, filters resolved. Joins implied by the model, not the SQL.

< 200ms
§ 03 · Render

Chart + drill

Visualization renders with drill paths attached at every aggregation level.

< 500ms
§ 04 · Source

To the truck roll

Drill from aggregate to job to crew to truck — without leaving the chart.

< 1s/level
Policy you can read

Metrics defined once.

Margin, NRR, gross retention, deal velocity — defined once in the semantic layer, queried everywhere. Disagreements about 'what counts as ARR' end on day one of implementation.

When the metric changes, every dashboard updates. Versioning preserves history. Auditable.

metrics · semantic.alm
# Semantic Metrics · canonical definitions
metric "net-revenue-retention":
  numerator   = arr.end_period(cohort)
  denominator = arr.start_period(cohort)
  format      = percent
  cohort      = customers.fixed_at(period.start)

metric "gross-margin":
  numerator   = revenue.recognized - cost.cogs
  denominator = revenue.recognized

  # cogs definition is itself a metric
  cost.cogs = labor.actual
            + materials.consumed
            + freight.allocated
            + equipment.utilization

metric "deal-velocity":
  expression = won.acv / cycle.days × 30
  segment_by = [region, segment, ae]
Where it lives

Mirrors to your BI of record.

BI

Looker/Tableau/etc

Allometry's semantic layer mirrors out — your BI tools query the same cube.

Warehouse

Snowflake/BQ

Cube backed by your warehouse; no data leaves your environment.

Notebook

Python/R/SQL

Direct SQL access for analysts. Same metrics, same definitions.

Embed

In-app insight

Embeds into CRM, ERP, Allometry — insight at the point of decision.

Real outcomes

"Our QBR used to take six analysts two weeks. Now it assembles itself. The CFO asked us last quarter to 'drill into the Atlanta margin' — answered in thirty seconds, on the call."

Naomi Bryce CFO · Operator D
−93%QBR prep time
28Cross-loop dimensions
1Source of truth
Operator D · executive QBR $340M ARR · 4 loops
See it on your data

Build your real cube.

Pick the question your team hates answering. We'll wire the cube, return the drill path, and run the QBR question in front of you live.