Speak spreadsheet
Use familiar names such as SUMIFS, IF, DAYS, CONCAT, and ROUND so the first step feels known.
Product concept · Collections administration
Excel-like answers, without leaving Collections.
Expression Studio gives administrators and consultants a visual way to create customer and invoice values, then turns each governed expression into SQL that runs where the data already lives.
1 native journeyFrom business question to tested field
30 + 30 fieldsReference fields at both data levels
Zero raw SQLFor admins building expressions
The problem
A simple SUMIF should not require a data journey.
Today, a request such as “sum open amounts for selected document types at customer level” usually follows one of two paths: move the data into LiveCube, or ask engineering to add or change a Java agent. The calculation is simple; both delivery paths add avoidable dependency and time.
Two current paths
Option 1 · LiveCube
- Move or replicate data out
- Rebuild logic in another model
- Return the result to Collections
Option 2 · Custom code
- Raise an engineering request
- Add or change a Java agent
- Test, release, and maintain it
Every new expression creates another integration or code-change cycle.
Native path
- Describe the value visually
- Translate from Excel semantics to SQL
- Simulate on selected customers
- Review governed SQL and parameters
- Publish as a native calculated field
One modular product journey, operating on the database and context users already trust.
Representative use case
Turn recurring customer requests into reusable fields.
The Costco request is one concrete example: sum open amount when document type is RV or DZ and status is Open. The same visual pattern can serve similar requirements across customers without another integration or agent change.
“Show the sum of open amounts for selected document types, for every customer.”
Selected document type open amountCustomer-level · Amount
For each Customer use SUMIFS to sum values in Open amount where Document type is RV or DZ and Status is Open
Costco Wholesale2 matching invoices$203,000
Experience principles
Familiar enough to start. Governed enough to scale.
The interface is not a thin SQL editor and it is not a spreadsheet clone. It uses spreadsheet language to make intent familiar, then adds product safeguards, simulation, lineage, and control.
Build visually
Compose a readable business sentence from approved fields, operators, and functions.
Simulate narrowly
Test on representative customers and their invoices before spending effort on a full run.
Reveal the execution
Show parameterized SQL, inferred result type, scope, and matched records without asking users to write SQL.
Capability boundary
Cover the SQL-translatable core of Excel first.
The product promise should be precise: support spreadsheet functions whose behavior can be represented safely and predictably in SQL at customer or invoice grain. Functions that depend on worksheet layout, volatile state, macros, or arbitrary code remain out of scope.
FamilySpreadsheet languageSQL executionCoverage
Conditional aggregationSUMIFS, COUNTIFS, AVERAGEIFSFILTER, CASE, SUM, COUNT, AVGPoC
MathSUM, AVERAGE, DIFFERENCE, PRODUCT, MIN, MAX, ROUND, ABS, percentagesSame-level numeric expressionsPoC
LogicIF, IFS, IFERRORCASE, COALESCEPoC
DatesDAYS, DATEDIF, EOMONTHDate arithmetic, DATE_TRUNCPoC
TextCONCAT, UPPER, LOWER, TRIMString functionsPoC
LookupsXLOOKUP, VLOOKUPGoverned joinsNext
Technical concept
Store intent, compile safely, execute natively.
The durable artifact is a typed expression definition, not a raw SQL string. A Java or Python service validates its fields and functions, compiles parameterized SQL for the supported database dialect, executes against the required grain, and records lineage.
Expression definitionTyped function, fields, scope, conditions
Policy validatorAllowlist, type checks, cost guardrails
SQL compilerParameterized, dialect-aware query plan
Native fieldCustomer or invoice result with lineage
No arbitrary SQL input
Sample and cost limits
Versioned expressions
Permissions and audit history
Product roadmap
Prove value quickly, then earn breadth.
The first release should solve high-frequency collections questions with exceptional clarity. Breadth follows observed demand, not a promise to reproduce every corner of Excel on day one.
Conditional calculations
- Customer and invoice output levels
- Math, logic, date, and text foundations
- CSV upload with 30 reference fields per level
- Selected-customer simulation and SQL preview
Govern and publish
- Role-based authoring and approvals
- Expression versions and dependencies
- Query cost analysis and runtime monitoring
- Calculated fields available across Collections
Advanced composition
- Governed lookups and reusable sub-expressions
- More SQL dialects and function mappings
- AI-assisted formula drafting with review
- Usage analytics and template marketplace
< 5 minMedian time from business question to validated expression
80%Target share of simple LiveCube or custom-code requests handled natively
100%Published expressions with owner, version, scope, and SQL lineage