login_FILL0_wght400_GRAD0_opsz24 MYEURESA
browse all use cases
use cases

Closed Loop Assistant - AI‑Supported Customer Feedback Response Process

  • by DEVK (Germany)

Objectives

  • Collect feedback after customer events (e.g., claims, new policy).
  • AI analyses free text and structured answers.
  • AI generates personalised responses with service or sales relevance.
  • Negative feedback triggers clear To Dos for internal teams.
  • Goal: improve satisfaction and identify sales opportunities

Progress status

Under development
  • Pilot phase – Manual testing currently ongoing.
  • Full automation planned upon successful evaluation.

Work team

  • Developed with an external partner
  • Project Management: Project Mgt: Market & customer Research Dept
  • External partner/team: External Partner: Hayuno (AI provider)

Work carried out to date

  • Regular non anonymous surveys sent after key events.
  • AI processes feedback and suggests suitable responses.
  • Developed response categories (positive, sales oriented, negative with To Dos).
  • Pilot process: feedback → AI → manual review → manual email.
  • Long term goal: automated responses and automated task routing.

First Results & Lessons learnt

  • The AI generates very good, highly personalised responses.
  • In some cases, AI responses are too enthusiastic or too emotional → needs fine tuning.
  • We need to train the AI on “DEVK language” (tone, style, level of formality, compliance).
  • The manual pilot process helps us understand which response types work best before automation.
  • No customer outreach yet; customer insights will follow in the next few weeks.

Difficulties encountered & Remainng Challenges

  • Technical integration needed for full automation.
  • Manual work slows process.
  • Ensuring consistent response quality