Designing the statement analysis and BI platform that gives lenders
deep cashflow visibility to make faster, more accurate credit
decisions.
My Role
Senior UX Designer Lead
Company
Indicina · Remote, Nigeria
Timeline
Apr 2023 – Present
Deliverables
Research · Dashboard Design · Mobile · Design System
Decide — Single Business Analysis: cashflow trends, income scoring
and behavioural insights
50%Engagement boost post-launch
90%+User satisfaction rating
Q1All metrics hit in first quarter
Overview
What is Decide?
Decide is Indicina's statement analysis and credit analytics
engine. It ingests raw bank statement data and surfaces structured
cashflow intelligence — income patterns, spending behaviour, debt
obligations — to help lenders approve the right customers faster.
When assigned to Decide, lenders were receiving raw data outputs
they couldn't interpret quickly. The interface presented numbers
without context.
My mandate was to transform raw financial data into clear,
actionable credit intelligence that any loan officer could act
on confidently.
Credit analysts, underwriters, and lending managers needing
cashflow visibility
Problem
Data-rich but insight-poor
Decide processed thousands of transactions per statement, but the
output screens presented raw figures with no narrative. Analysts
had to manually calculate ratios and patterns. Lenders were making
decisions based on incomplete mental models.
"I can see the numbers but I don't know what they mean for the
loan decision. I end up second-guessing myself on every
application."
Unstructured Data
Transaction data was displayed as raw lists with no grouping,
categorisation or trend context.
Manual Analysis Burden
Analysts spent 20+ minutes per statement building their own
Excel summaries to extract meaning.
No Comparative Context
Income figures had no benchmark — analysts had no way to gauge
whether metrics were good or bad for a given loan size.
Research
Understanding how analysts actually think
I shadowed credit analysts during live statement reviews, running
contextual inquiry sessions to observe their mental models — what
they looked for first, what caused hesitation, and what prompted a
confident decision.
NC
Ngozi Chukwu
Senior Credit Analyst
"The first thing I want to know is: does this person earn
consistently? Everything else flows from that. But the current
tool makes me calculate it myself from raw figures."
Pattern-seekingRisk-awareDetail-oriented
Key findings: The first mental model analysts
built was income consistency — yet it required 4 manual
calculations. 73% of analysts said they "weren't sure" whether a
statement was strong or weak without their own Excel model.
Decision confidence was the primary unmet need, not raw data
access.
Process
Designing for analyst intuition
01
Mental Model Mapping
Mapped the analyst's decision journey — what signals they
sought, in what order — to create a hierarchy of information
that matched how experts actually think.
02
Data Narrative Framework
Designed a "cashflow story" framework — leading with the
most decision-relevant signal (income consistency), then
layering in expenses, obligations and behavioural patterns.
03
Automated Intelligence Layer
Worked with data science team to surface pre-calculated
ratios — DTI, income consistency score, spending volatility
— so analysts read insights, not raw numbers.
04
Prototype & Validate
Ran 3 rounds of prototype testing with active credit
analysts. Iterated on the health score visualisation, tab
structure, and summary card hierarchy based on direct
feedback.
Solution
Turning raw transactions into credit intelligence
The redesigned Decide platform presents financial data as a
structured narrative — leading with the most decision-relevant
signals, contextualising every metric, and guiding analysts to a
confident recommendation.
Decide — Business Analysis showing monthly cashflow trend,
income vs expense breakdown and health scoring
📊
Cashflow Narrative View
Income, expenses and net cashflow displayed as a monthly trend
story — not a table of numbers.
⚡
Automated Ratio Calculations
DTI, consistency score, and income stability calculated and
displayed automatically — removing mental arithmetic from the
workflow.
🔍
Behavioural Signals
A dedicated Behavioural tab surfaces spending patterns,
recurring obligations and anomalies invisible in raw figures.
💡
Business Health Score
A single composite score combining revenue strength, expense
control and consistency — an instant entry point before deep
review.
Outcomes
Analytics that drive decisions
50%Engagement boost post-launch
90%+User satisfaction rating
Q1All metrics hit in first quarter
"For the first time I can look at a Decide report and know exactly
what decision to make without downloading anything or building my
own spreadsheet."
— Credit Analyst, Indicina partner lender
Learnings
What this project taught me
Data without narrative is noise. The biggest
design win wasn't a new feature — it was restructuring existing
data into a story analysts could read in 60 seconds.
Expert users have deeply ingrained mental models.
Designing for analysts means matching their existing cognitive
flow, not imposing a new one.
Confidence is a design outcome. "User
satisfaction" in this context meant analysts trusted the tool
enough to make a final decision without a second source.
Collaborate with data science early. The best
design decisions came from understanding what the algorithm
could calculate, not just what users wanted to see.
