PURPLE CARE

Purple Care

One complaint. A connected path to better care.

An integrated complaint-intelligence solution connecting routing, structured analysis, emotion and response review, FAQ improvements and message quality.

Purple Care brings Minwon Ieum's knowledge-graph routing and Dong-A University's HF CARE AI complaint-analysis research into one solution offering. Each capability retains its development or research status; live end-to-end integration requires project scoping and validation.

Try Purple Care demo Discuss implementation

01
Route with evidence
02
Structure and analyze
03
Improve responses
04
Staff review

Five capabilities, one solution

Explore each role and its sample demo within Purple Care.

In development

Complaint Routing Agent

Matches complaint content and location to duties, recommending responsible teams with supporting evidence.

Input: Complaint text, optional location and administrative knowledge graph
Output: Recommended department, team and staff; related duties, jurisdiction, keywords and graph relationships

Try this agent

Research outcome

VOC Structuring Agent

Structures complaints into facts, original goals, appraisal and current requests.

Input: VOC text and business taxonomy
Output: Four core fields and five classification axes: business, product, stage, issue and situation

Try this agent

Research outcome

Emotion & Response Analysis Agent

Compares expressed emotions with response characteristics to identify service improvements.

Input: VOC, responses and satisfaction feedback
Output: Comparisons of direct answers, request outcomes and alternative guidance by emotion

Try this agent

Research outcome

Recurring VOC & FAQ Analysis Agent

Groups recurring requests and compares them with existing FAQs to find coverage gaps.

Input: VOC records and existing FAQs
Output: Recurring request clusters, FAQ coverage and improvement candidates

Try this agent

Review function implemented

Message Quality Review Agent

Checks outbound messages against a rubric and proposes revisions for staff review.

Input: Draft messages and a quality rubric
Output: Checks for purpose, body, submission channel, considerate wording, closing and greeting; gaps and revision suggestions

Try this agent

Explore all demos

One complaint. A connected path to better care.

An integrated complaint-intelligence solution connecting routing, structured analysis, emotion and response review, FAQ improvements and message quality.

Five capabilities, one solution

Explore each role and its sample demo within Purple Care.

In development

Complaint Routing Agent

Matches complaint content and location to duties, recommending responsible teams with supporting evidence.

Input: Complaint text, optional location and administrative knowledge graph
Output: Recommended department, team and staff; related duties, jurisdiction, keywords and graph relationships

Try this agent

Research outcome

VOC Structuring Agent

Structures complaints into facts, original goals, appraisal and current requests.

Input: VOC text and business taxonomy
Output: Four core fields and five classification axes: business, product, stage, issue and situation

Try this agent

Research outcome

Emotion & Response Analysis Agent

Compares expressed emotions with response characteristics to identify service improvements.

Input: VOC, responses and satisfaction feedback
Output: Comparisons of direct answers, request outcomes and alternative guidance by emotion

Try this agent

Research outcome

Recurring VOC & FAQ Analysis Agent

Groups recurring requests and compares them with existing FAQs to find coverage gaps.

Input: VOC records and existing FAQs
Output: Recurring request clusters, FAQ coverage and improvement candidates

Try this agent

Review function implemented

Message Quality Review Agent

Checks outbound messages against a rubric and proposes revisions for staff review.

Input: Draft messages and a quality rubric
Output: Checks for purpose, body, submission channel, considerate wording, closing and greeting; gaps and revision suggestions

Try this agent