tenet.app/catalog
1
Discover
2
Review
3
Catalog
4
Detail
🔍 Search metrics…
All Growth Finance Product AI / ML
Monthly Recurring RevenueMRR
Sarah ChenPublished
Daily Active UsersDAU
Marcus JohnsonPublished
Net Revenue RetentionNRR
Tom NguyenPublished
Customer Acquisition CostCAC
Priya PatelDraft
AI Cost Per InferenceAICPI
Claude ShannonPublished
Churn RateCHURN
David KimPublished
Token ThroughputTPUT
Claude ShannonPublished
$ tenet discover --source dbt --source snowflake
✓ Connected to Snowflake (WJSHYUS-ZAB77981)
  Found 28 tables in ANALYTICS schema
✓ Parsing dbt manifest (v1.7)
  Found 23 models: 9 source → 9 staging → 5 intermediate → 9 mart
✓ Analyzed query history (30k queries, 14 users, 90 days)
✓ AI inference on discovered metadata
✓ Attributed owners from query patterns (7 owners, 5 teams)
Discovery complete — 7 metrics found
MRR · 94% DAU · 91% NRR · 89% Churn · 87% AI Cost · 86% TPUT · 83% CAC · 78%
Generated 7 KPI definitions → opening review
Review & Publish
7 to review
✓
Monthly Recurring RevenueSum of recurring subscription revenue, monthly normalized
94%Draft
✓
Daily Active UsersUnique users with ≥1 meaningful action in 24h
91%Draft
✓
Net Revenue RetentionRevenue retained from existing customers incl. expansion
89%Draft
✓
Churn RatePercentage of customers lost in the period
87%Draft
✓
AI Cost Per InferenceAvg cost per model inference incl. compute & API fees
86%Draft
✓
Customer Acquisition CostTotal S&M spend ÷ new customers acquired
78%Draft
✓
Token ThroughputTokens processed per second across all models
83%Draft
☐ Select All0 of 7 selected
Publish Selected
🔍Search metrics…
AllGrowthFinanceProductAI / ML
Monthly Recurring RevenueMRR
Sarah ChenPublished
Daily Active UsersDAU
Marcus JohnsonPublished
Net Revenue RetentionNRR
Tom NguyenPublished
Churn RateCHURN
David KimPublished
AI Cost Per InferenceAICPI
Claude ShannonPublished
Token ThroughputTPUT
Claude ShannonPublished
Customer Acquisition CostCAC
Priya PatelDraft
Showing 7 of 24 metricsView all →
Catalog / AI Cost Per Inference
AI / ML OpsPublished✓ Trusted
AI Cost Per InferenceAICPI
OverviewFormula & SourcesStakeholdersReportingHistory
Definition
The average cost per model inference call, including compute, API fees (Claude/OpenAI), and infrastructure overhead.
Available via
Slack
Natural language queries in any channel
"What is AICPI?"
M
MCP
AI agents via Model Context Protocol
tenet.get_metric("aicpi")
Why This Matters
The primary cost efficiency metric for AI-native products, monitored by ML engineering and finance teams. Determines whether inference workloads are cost-sustainable and drives model optimization decisions.
Owner
CS
Claude Shannon · AI Engineering Lead