Selected work
Product design case study · 2025

Close IQ

A real-time copilot that rides a live client call — designed around restraint. The hard problem wasn’t surfacing help. It was building a system whose primary job is staying silent, so the one cue it does give is trusted, not ignored.

PolicyAdvisorProduct designReal-time UXRestraint & trustAI product
Client

PolicyAdvisor

Role

Designer + build (design & front-end)

Approach

Designer-first · restraint-led

Status

Shipped · in production

The silenceThe signalThe trustThe one cue The silenceThe signalThe trustThe one cue
The scenario everything was designed around
An advisor is on a live call. A copilot that bombards them is the thing that loses it.

There are regulatory beats that must be hit — disclose pre-existing conditions, confirm the stability period, give the suitability rationale, explain exclusions, confirm limits — and a real client on the line who needs to feel heard. The instant the advisor is reading a wall of AI notifications, they’ve stopped listening, and the tool has become the problem.

So I didn’t design a dashboard of suggestions. I designed restraint — a context engine whose primary job is deciding what not to say, so the rare cue it does surface is the one the advisor trusts and acts on.

Feature 01 · silent by default

The best thing it can do is shut up.

A live-call assistant that talks too much is the thing that loses the call. So the resting state is — nothing.

Silence is the defaultNo cue is the normal state, not a gap.
It trusts the advisorAssumes competence until proven otherwise.
Always watchingTracking everything — surfacing almost none.
Earns every interruptA cue has to beat a high bar to appear.
Wireframe — structural sketch, not UI
ADVISOR’S SCREEN — their CRM, their call, their flowCLOSE IQ — restinglistening · 0 cuessilent on purpose95% of the screen is the advisor’s. The tool is a quietcorner that stays quiet — interruption is the exception.

The resting state isn’t an empty dashboard waiting to be filled — it’s a deliberate silence. The advisor runs the call; the tool earns the right to speak.

Close IQ · the resting state · production UI
crm.policyadvisor.com · Daniel R. — Term Life application
Close IQon call
2:18End call
Listening. Nothing to surface.The call is going well. Close IQ is tracking everything — and saying nothing, on purpose.
0 cues shown · 14 evaluated & suppressed
Powered by PolicyAdvisor

Most copilots are designed around what they can show. This one is designed around what it refuses to.

Feature 02 · the context engine

Its real job is deciding what not to say.

The engine tracks everything — compliance beats, client signals, what’s been covered. Then it suppresses almost all of it.

Everything inCompliance, client cues, call state.
Ranked & dedupedStakes, timing, already-covered.
Most of it droppedSuppression is the primary function.
One thing outOnly what clears the bar surfaces.
Wireframe — structural sketch, not UI
ALL SIGNALScompliance beatsclient hesitationmissed stepsupsell, tone, pace~14 / minuteSUPPRESSalready covered?low stakes?would interrupt?drop ~95%ONE CUEhighest-stakes, well-timedsurfaced — once

The funnel is the product: many signals in, a suppression stage that drops almost all of them, exactly one cue out. The engineering exists to stay quiet.

Close IQ · the context engine · production UI
crm.policyadvisor.com · Daniel R. — Term Life application
Close IQon call
4:41End call
High · surfaced

Give the suitability rationale — explain why this plan fits Daniel’s specific needs before moving to close.

timed to a natural pause · not mid-sentence
9 other signals evaluated this minute — suppressed: already covered, low stakes, or would interrupt.
Powered by PolicyAdvisor

A context system that surfaces everything is just noise with a model attached. The intelligence is in the suppression.

Feature 03 · one cue, ranked

One at a time. Never stacked.

When it does speak, it’s a single prioritised nudge — glanceable in the half-second an advisor can spare.

Priority-rankedHigh before Medium — one shows.
Never a stackA second cue waits; it never piles on.
Timed to a pauseSurfaces between beats, not mid-sentence.
Clears itselfCovered → it disappears, no dismissing.
Wireframe — structural sketch, not UI
ONE CUE — visiblehighest priority onlyglanceable · ~0.5sauto-clears when coveredNEXT CUE — held backwaits for the slot to freeLOWER — may never showresolved silently if covered

A strict queue of one. The next cue does not appear until the current one clears — the advisor is never reading a list while a client is talking.

Close IQ · one cue at a time · production UI
crm.policyadvisor.com · Daniel R. — Term Life application
Close IQon call
5:02End call
High

Walk through the key policy exclusions so Daniel understands what is and isn’t covered — before he commits.

one cue · never stacked · clears when covered
A Medium cue is waiting — it will not show until this one resolves.
Powered by PolicyAdvisor

Two cues at once means the advisor reads instead of listens. The one-at-a-time rule is the whole point.

Feature 04 · the ambient checklist

Compliance that tracks itself.

Every regulatory beat is watched and checked off as the advisor naturally covers it — nothing to manage mid-call.

Auto-checksCovered in conversation → ticked.
Ambient, not interactiveNothing to tap, nothing to manage.
Never nags covered itemsDone is done — it goes quiet.
Nothing regulatory slipsThe miss is what it watches for.
Wireframe — structural sketch, not UI
ADVISOR TALKS NATURALLY — no checklist interactionLISTEN + MATCHbeat covered in speech?→ tick it, silentlyCHECKLISTfills itself · 4/10only the GAP can raise a cue

The advisor never touches the checklist — it fills from the conversation. It’s ambient by design: present when glanced at, silent otherwise. Only an un-covered required item can escalate.

Close IQ · the ambient checklist · production UI
crm.policyadvisor.com · Daniel R. — Term Life application
Close IQon call
3:55End call
Call checklist4 / 10 · auto-tracked
Pre-existing conditions disclosed
Stability period confirmed
Suitability rationale given
Exclusions explained
Coverage limits confirmed
+ 5 more · tracked silently in the background
Powered by PolicyAdvisor

A checklist the advisor has to manage mid-call is another thing stealing attention from the client. So they don’t manage it.

Feature 05 · it trusts the advisor

It earns the one interruption.

It assumes competence. It doesn’t nag, repeat, or escalate — it stays silent so that the one time it speaks up, the advisor listens.

Assumes competenceThe advisor runs the call, not the tool.
No nagging, no escalationIt never raises its voice.
Breaks silence only to prevent a real missA required disclosure about to be skipped.
Close IQ · trust, and the one intervention · production UI
crm.policyadvisor.com · Daniel R. — Term Life application
Close IQon call
6:30End call
The one time it interrupts

Daniel’s about to commit — and exclusions haven’t been explained. This is a required disclosure. Cover it before you take the close.

Not a nag. Not a tone change. The single intervention it was saving its silence for.
Eleven lower-value moments this call were watched and let go — because trusting the advisor on the small things is what makes this interruption land.
Powered by PolicyAdvisor

Trust isn’t a feature you add. It’s the silence you keep — so the one cue that matters is the one they act on.

The design spine — suppression, visualised

The product is the funnel.

Everything the engine does serves one outcome: many signals in, almost nothing out.

~14
signals evaluated per minute
~95%
suppressed: covered, low-stakes, or intrusive
1
cue surfaced — ranked, timed, trusted
The outcome
A copilot the advisor forgets is there — until the moment it matters.

Designing from the bombardment failure instead of the feature list set what every screen had to do: stay silent by default, suppress 95% of what it tracks, surface one ranked cue at a natural pause, track compliance without asking the advisor to manage it, and break silence only to prevent a real miss. The restraint pattern — trusting the user enough to say nothing — is the spine of the whole PolicyAdvisor AI suite.

Reflection

The hardest thing to design was everything it doesn’t do.

A live-call assistant is the rare product where the obvious version — helpful, present, surfacing insight — is the one that fails. Every notification is a moment the advisor isn’t listening to a human being who’s deciding whether to trust them.

The decision I’m proudest of has no UI: the resting state is silence, and the engine’s primary function is suppression. It would have demoed better with more on screen. It works better with almost nothing — because trust is the silence you keep, so the one cue that matters is the one they act on.