Ultra PatternsAI interface patterns with receiptsULTRA project

The library

19 patterns, in three groups

Pick the group that matches what you are building. Every card shows what the pattern is, when it fits, the words on screen and how to build it. Where a test has run, the receipt shows what happened.

EvidenceTested0Tested (pilot)3Seen in a case16On the list15What the labels mean
Interface type
Evidence
More filters: design goal and pattern family
Design goal
Pattern familyWhat the families mean
Showing 19 of 19 patterns. Within a row, chips add up; across rows, they narrow down. Counts show what each chip would leave.

Chat or assistant

Conversations, copilots and guided dialogues: how the assistant asks, answers, refuses and hands over.

10 of 10
Pattern 04 · Chat or assistant, Dashboard with AI insightsClassification with confidence and reasonA label, how sure the system is, and the evidence it used, so a person can check the judgement instead of reading a bare percentage.Seen in a casePattern 06 · Chat or assistant, Dashboard with AI insightsLow-confidence hand-off to a personBelow a threshold the AI stops, says what it could not determine, and hands the item to a person instead of guessing.Seen in a casePattern 09 · Chat or assistantMatch recipeA persistent five-signal checklist updates while the user writes a brief in free text, without forcing a form.Seen in a casePattern 10 · Chat or assistantWhy this questionWhen an assistant asks a follow-up, it says in one line why the answer will improve the result.Seen in a casePattern 11 · Chat or assistantStaged conversation: the code decidesThe conversation always sits in one named phase. The model can only signal readiness; the app decides when to move on, refuse or repeat.Seen in a casePattern 12 · Chat or assistantOne question per turnEach assistant turn asks exactly one thing. People answer without juggling, and the answer format makes the rule enforceable.Seen in a casePattern 13 · Chat or assistantOut-of-scope refusal with a next stepWhen a request leaves the assistant's remit it says so, offers a safe next step and, where lives may be at stake, a helpline. Decided by code, not by mood.Seen in a casePattern 14 · Chat or assistantParaphrase before you adviseThe assistant first mirrors what the person said in its own words, then asks or suggests. Only what was said, never what it guessed.Seen in a casePattern 15 · Chat or assistantShow an example, not a ruleEvery tip comes with a sentence the person could actually say. In both Goose rounds, the example dialogue was the most valued feature.Seen in a casePattern 16 · Chat or assistantYour words first, then feedbackIn practice mode the person writes the reply; the assistant gives feedback on it instead of writing it. The skill stays with the person.Seen in a case

Dashboard with AI insights

Lists, labels and scores produced by AI: how people see what the system judged and why.

6 of 6
Pattern 17 · Dashboard with AI insightsBoard first, guided brief on requestThe person lands on a board they order themselves; the AI's one-task-at-a-time brief opens per task, on request. Five of seven accountants preferred it.Tested (pilot) · 7 participantsPattern 05 · Dashboard with AI insights, Chat or assistantUrgency labels for prioritisationA small set of urgency levels on every item, so a full inbox can be scanned in seconds and the order of work is explicit.Seen in a casePattern 07 · Dashboard with AI insights, Chat or assistantWhy is this urgentA one-line reason behind an AI judgement, available on click rather than always on screen.Seen in a casePattern 08 · Dashboard with AI insightsReview a colleague's work checkpointAI-prepared work by a junior reaches a senior with the source, the junior's note and two clear actions.Seen in a casePattern 18 · Dashboard with AI insights, Chat or assistantOne proposed action, one named buttonOne AI proposal at a time, with its verb on the button: Send this reply. Two proposals side by side: five of seven completed the task. One: all seven.Seen in a casePattern 19 · Dashboard with AI insightsTeam workload view with availabilityEach colleague's open and overdue tasks and whether they are in, behind a navigation item with a text label. Valued by all seven, once found.Seen in a case

An app that adapts

Forms and flows that fill in or adapt: how to keep the person the author.

3 of 3
Pattern 01 · An app that adaptsOn-demand suggestions per fieldFields stay empty until the user asks for options, so the user chooses a direction instead of hunting for errors.Tested (pilot) · 6 participantsPattern 02 · An app that adaptsPre-filled form with reviewThe AI fills every field first and the user checks. Suited to facts the system can verify, risky for judgement content.Tested (pilot) · 6 participantsPattern 03 · An app that adapts, Chat or assistantPrivacy check before sending personal dataBefore text goes to a model, the app spots personal data and lets the user send, strip or skip the AI step.Seen in a case

On the list

Which one should we test next?

Candidates from the libraries we build on and from the cases. Vote by e-mail; each link opens a message with the name filled in.

Offer your product for the next test