ALLY PROACTIVE AI · UX CASE STUDY
A UX research case study on how investment bankers assess, verify, and act on AI-surfaced acquisition signals during high-stakes deal work.
ROLE
UX Researcher
RESEARCH FOCUS
Trust, explainability, and decision-making
METHODS
Landscape review · journey mapping · concept testing
03 / 17 · THE REAL PROBLEM
Three research questions behind a simple-sounding brief.
I translated the broad question of AI trust into three research questions that shaped the product direction: what evidence bankers need, how they triage signals, and what lets them act with confidence.
01
The trust problem
Bankers are paid to verify everything. Output must be sourced, weighted, and easy to push back on.
02
The attention problem
A banker reads an insight headline in roughly 8 seconds. Thesis, confidence, urgency, and sources must be clear at a glance.
03
The action problem
Reading is expensive. Acting should be cheap: one click to pipeline, share, or request analysis — provenance included.
04 / 17 · PROCESS
A research-to-design approach — explore, focus, test, and refine.
Discover broadly, define narrowly. Develop broadly again, deliver one clear direction. The framework forces honest choices about what to leave out.
01
Discover
Understand the space: research, competing tools, explainability patterns.
02
Define
Frame the right problem: journey, information architecture, design principles.
03
Develop
Explore solutions: confidence models, trust interfaces, wireframes, flow tests.
04
Deliver
Ship one direction: four connected screens, a token system, transparent edge cases.
07 / 17 · DESIGN PRINCIPLES
Research insights became three design principles.
01
The AI is a colleague, not an oracle.
Treat output like a junior analyst’s memo: confident, defensible, and easy to challenge.
02
Earn the click in 8 seconds.
Surface thesis, confidence, and freshness on the card. Full explainability lives one click deep — not zero, not two.
03
Make the decision the easy part.
One click to pipeline, share, or queue analysis — each carrying its origin and supporting evidence automatically.
05 / 17 · DISCOVER — FINDINGS
What investment bankers told me they struggle with.
Trust · “Why this, why now?”
Alerts arrive with no reasoning chain. Analysts re-do the research to confirm — defeating the purpose.
Volume · Alert fatigue
Tools fire dozens of low-relevance alerts daily. Analysts learn to mute — then miss the important one.
Context · Isolated insights
An insight that doesn’t connect to a mandate or pipeline feels orphaned and hard to action.
Friction · Citation-hygiene tax
Sharing with an MD means manually copying sources and rebuilding the reasoning by hand.
10–13 / 17 · PRODUCT EXPERIENCE
Following a signal from first glance to a confident decision.
I mapped one analyst’s morning workflow across four connected moments, using each step to test whether the product gave enough evidence to move forward with confidence.
01 · DASHBOARD · 07:42
What’s worth my attention today?
$2.4B
Pipeline value
11.5h
Analyst hours saved
02 · AI SIGNALS · 07:43
Mosaic Industries may be seeking acquisition.
87%
High confidence · rising · surfaced 2h ago
03 · INSIGHT PANEL · 07:44
Conviction comes from precedent, not a bare percent.
73%
Pattern accuracy
2 flags
Counter evidence
04 · PIPELINE · DECIDED
The decision lands — and the origin travels with it.
IDENTIFIED
QUALIFYING
PITCHING
AI Insight · Supporting signals attached · Risk: sector volatility
14–16 / 17 · BUILDING THE PROMISE
Making research findings actionable for product and engineering.
I translated research insights into clear interaction states, evidence requirements, and edge cases so the team could preserve user trust through implementation.
Every variant & state — specified, not assumed.
High, medium, and low confidence. Rising, stable, and falling trajectory. Default, hover, focus, loading, empty, and error states.
Where the design has to keep its promises.
Low-confidence signals are demoted, not hidden. Stalled confidence becomes information. Disagreement is captured as useful feedback.
17 / 17 · REFLECTION
What I learned as a UX researcher working with AI products.
Trust in AI is not created by a single confidence score. My research showed that people need visible reasoning, meaningful counter-evidence, and a clear path to make accountable decisions.