Context
Market context arrives scattered: one tab for prices, a spreadsheet for indicators, a screenshot pasted into a note. Reading a series and its history means reassembling the same picture every session. Paper and historical context only; no trade advice.
The challenge
Without a product layer, context gets rebuilt for every question. Prices in one place, historical series in another, indicators kept in a sheet, a screenshot pasted into a note as the record of what was looked at. Nothing holds the state of a review, so returning to the same series a week later means reconstructing the view from memory and comparing periods by eye.
Our work
A data product organised around separated layers: view · analyse · history · the user's own decisions. Prices, indicators and market context render in one reading interface. Historical series sit beside current state instead of living in an export. The user's own decisions and inputs are kept as a distinct layer, never merged into the underlying data. Interface behaviour is treated as product logic, not decoration.
What was built
- Research interface
- Prices, indicators and market context presented as one readable surface instead of tables reassembled for every question.
- Historical context
- Historical series sit next to current state, so a period can be examined without exporting the data into a separate tool.
- Separated layers
- Viewing, analysis, history and the user's own decisions are kept as distinct layers, so recorded data is never confused with judgement.
- Decision boundary
- The product presents context and stops there: no signals, no recommendations, no forecasts and no trade advice.
How the process looks
Before
- Prices, series and indicators live in separate places
- Every review is rebuilt from memory
With the system
- View
- Analyse
- History
- User decisions
The evidence
No system captures have been cleared for publication yet. This page describes what was built and how it works, without passing off a mock-up as a real screen.