Use case / Finance copilots
Inspect entity, metric, period, role and source snapshot alongside the tool call and answer. Verify derived numbers through their operands and formula.
Tensile / Finance copilot
Bind
Entity, metric, requested periods, reporting snapshot and acting role.
Retrieve
Retain the source values and distinguish provider errors from model errors.
Derive
Inspect (120 − 100) ÷ 100, sign, units and rounding.
Illustrative records · proposed experience
Explain Q3 revenue growth from the correct snapshot
Example definition · origin I-104 · draft v1
Check C-104 / v1 · Bind the requested entity and period
check: C-104/v1 condition: Bind the requested entity and period source_fixture: finance-v1 evaluation: E-104/v1 method: independent scoped evidence assertion missing_required_source: Unknown changed_source_or_policy: new qualification required activation: disabled in this concept
Check C-105 / v1 · Use the pinned source snapshot
check: C-105/v1 condition: Use the pinned source snapshot source_fixture: finance-v1 evaluation: E-104/v1 method: independent scoped evidence assertion missing_required_source: Unknown changed_source_or_policy: new qualification required activation: disabled in this concept
Check C-106 / v1 · Compute (120 − 100) ÷ 100 = 20%
check: C-106/v1 condition: Compute (120 − 100) ÷ 100 = 20% source_fixture: finance-v1 evaluation: E-104/v1 method: reproducible arithmetic missing_required_source: Unknown changed_source_or_policy: new qualification required activation: disabled in this concept
Check C-107 / v1 · Respect the acting role’s access
check: C-107/v1 condition: Respect the acting role’s access source_fixture: finance-v1 evaluation: E-104/v1 method: independent scoped evidence assertion missing_required_source: Unknown changed_source_or_policy: new qualification required activation: disabled in this concept
Pack groups Evaluations. Each Evaluation pins reusable Check versions and an explicit aggregation policy.
Clickable product concept · invented records · no external execution.
A closer look / illustrative case
Both source operands can be correct while the answer is wrong. With Q2 = 100 and Q3 = 120, dividing the change by Q3 produces 16.7%; the requested Q2 comparison is 20%.
Bind
Entity, metric, requested periods, reporting snapshot and acting role.
Retrieve
Retain the source values and distinguish provider errors from model errors.
Derive
Inspect (120 − 100) ÷ 100, sign, units and rounding.
Challenge
Wrong entity, stale snapshot, missing baseline and denied role.
Verify
Review the saved answer and any subsequent correction against the intended question.
An accurate calculation on the wrong entity or period still fails the customer’s question.
Context binding
Compare the user’s intended question with the selected entity, metric and reporting snapshot. Keep retrieval errors separate from model errors.
Explore the connected product ↗
Derivation
Retain source cells, operands, operator, denominator, sign and units. Test direct and derived answers under the applicable access role.
Explore the connected product ↗
Outcome
Measure correction and follow-up on representative work. A grounded number does not establish a useful or accepted financial decision.
Explore the connected product ↗
What your team gets
01
The question’s entity, metric, period, snapshot and access role.
02
Source operands, reproducible formula, units and checks.
03
A reviewed answer and linked corrections; useful decision impact must be measured.
Start with an affected customer job, the expected result and the evidence your team would accept.
Discuss a concrete workflow ↗
AI reliability, from production to customer outcome.