Capture · The AI interaction layer
One source of truth for every number — that your AI can actually use.
Capture's final output — the governed semantic layer your AI reads the business through.
"How many active members do we have?" shouldn't have three answers. Flash defines each business metric once, certifies it, and calculates it one way — so a dashboard, an API call, and an AI agent never disagree.
GMV
$198,862
Active members
bought in 90 days
Open rate
of delivered emails
Redemption rate
per coupon code
"Why did active members dip last month?"
→ uses the certified active members definition, breaks it down by store & channel — never invents a number.
The problem
The same metric means three different things in three tools.
Definitions drift
GMV, churn, redemption rate — each gets redefined in every report, so two teams quote two numbers and nobody knows which is right.
AI invents numbers
Point an AI at your raw data and it makes up its own calculations against fields it doesn't understand — confident, untraceable, and often wrong.
Custom metrics are risky
Letting analysts dig through raw data risks exposing the wrong customer's records — one mistyped filter away from a data incident.
How it works
A governed catalog, one calculation, and metrics AI can use safely.
80+ business metrics across seven categories, each with a single definition, plain-language guidance, and one way it's calculated — read by every tool from the same source.
Sales & Revenue
GMV, net GMV, orders, AOV, refund & discount rate.
Customer Value & Retention
Active members, retention, churn-risk, LTV signals.
Membership & Lifecycle
New / total / marketable members, tier moves, opt-in rate.
Points & Rewards
Issued, redeemed, expired, breakage, outstanding liability.
Email & Messaging
Sent, delivered, open / click / bounce / unsubscribe rate.
Web & Portal
Page views, sessions, portal blocks, UGC, abandoned cart.
Attribution & Growth
Referrals, ambassadors, D2C channel orders, coupons.
+ your own
Add your own metrics with a guided builder or safe, governed queries — test them before they go live.
One definition · every surface reads the same number
ONE DEFINITION
Active members
Defined and certified once, with a single owner. Any change goes through a formal review, so there's always an approval trail.
Dashboards
same number, same definition
Reports & API
same number, same definition
SoTag in Slack
same number, same definition
MCP AI agents
same number, same definition
Certified, owner-owned definitions
A certified metric has a single owner; any change to how it's defined goes through a formal review, so the definition has an approval trail instead of quietly drifting.
Custom metrics, safe by design
Build with a guided, no-code editor, or write governed, read-only queries that can only reach approved data — automatically kept isolated to your own team, not by a filter someone has to remember.
Ask in plain language
Ask a question; Flash picks the right governed metric, applies its certified definition, and breaks it down by store, channel or tier — grounded in real data, never guessed.
Why it's different
Dashboards let everyone redefine the number. Defining a metric once still doesn't make it AI-safe. Flash does both.
Dashboards let every report invent its own version of a metric — so the numbers drift. Tools that define metrics once still leave AI access and data isolation up to you. Flash governs the definition andmakes that same metric the only way AI agents and dashboards can read it.
Typical approach
Dashboards & spreadsheets
Every report redefines GMV or churn — three tools, three numbers.
Flash, by design
One certified definition with a single owner and full history — change it once, everywhere updates.
Typical approach
Metric-modeling tools
Define metrics once, but leave AI access and data isolation to you.
Flash, by design
The same governed metric is available to your AI as an MCP tool, with each team's data kept isolated automatically.
Typical approach
AI tools that query raw data
Write queries against data they don't understand — and make things up.
Flash, by design
AI can only use governed metrics; it returns the correct number or says it can't answer.
AI & innovation
Governed metrics are what make Flash's AI trustworthy.
SoTag in Slack and MCP agents don't query your raw data — they use the same governed metrics your dashboard reads. So the number an agent quotes is the number on your screen, every time.
Governed metrics, not raw data
Metrics are made available to AI agents as governed MCP tools — permission-limited, team-isolated, and always using the certified definition — so agents act on the same truth as your team.
One calculation, identical everywhere
Every built-in metric is calculated one way — so 'ask the AI' and 'open the dashboard' can't return different numbers.
Grounded, or it holds back
If a question doesn't match a governed metric, the AI says so instead of making up an answer — trust over a confident wrong number.
What changes for the business
One number per metric, an approval trail behind it, and AI answers you can actually trust.
80+
governed metrics across 7 categories
1 definition
per metric — certified, with a single owner
1 calculation
every dashboard, API & agent reads the same
0 guesswork
AI uses governed metrics, never your raw data
Every metric defined once, so dashboards, API and AI never disagree.