LG My Data POC Project

Overview

Life Style AI Assistant with My Data

With the revision of the Personal Information Protection Act, conditions have been established to provide personalized services using data with user consent. This is a preliminary research project aimed at proposing the K-Data project, a MyData pilot initiative by the Korean government.

Company

Role

  • Lead Product Designer

Deliverables

  • Service Strategy

  • Concept UX Design

  • Service Architecture

Project Date

  • FEB 2020 - AUG 2020
    6 Month

Context & Background

The key issue with existing services, such as AI chatbots, was their inability to leverage my data, resulting in rule-based services.

In the case of personalized product recommendation services, it was observed that user purchase conversion rates in the e-commerce industry increased by 50% compared to previous.

Hyphothesis : HMW Approach

“How might we safely utilize user data to understand their situations and provide a better life through our services?”

We conducted benchmarking by reviewing existing references on My Data and collaborated with key stakeholders, including the head of the data security team, the public sector account manager, and AI Lab engineers, to design the My Data service architecture. Following this, we conducted user interviews and stakeholder workshops with partner service companies to align on data provision and consortium formation. Based on these insights, I designed a POC concept that incorporated their requirements.

User Research

Stakeholders Interviews

  • Mana in charge of maternity policies at Seocho-gu Health Center

  • SSG.com, GS Retail and other Retail companies partnership manager

  • LG CNS Data & Security, LG CNS AI Lab team leader

  • LG CNS Public sector business Account Manager

  • K-Data Project Leader

Benchmark

  • Digi.me My data Personal Data Storage Benchmark

  • Google Assistant, Apple Health app Service Flow Benchmark

  • Amazon Web Service EC2, Google Cloud Platform Cloud Storage Cloud Data storage benchmark

  • On-premise NLU/NLP Service Benchmark

Outcomes

LG Mydata Concept(POC)

Pregnant Women’s Care My Data Service Concept for K-Data Pilot Project

Single Sign On - Simple On Boarding Process

User research revealed that most drop-offs occur during onboarding. To address this, we streamlined the process by clearly defining the required information at each step, allowing users to input their details and complete the process with the "Continue" button. For personal storage, we simplified the process by only requiring consent for a popular cloud service. Users without a cloud account can easily sign up by linking their existing registration, enabling seamless sign-on with a single ID.

Seamless Data Integration, Led by Me

Requiring users to log in again for each data connection can be a hassle. To simplify this, we leveraged the open banking authentication system to automatically list services linked to the user based on the information provided during registration. Users can then select which information to integrate via a layer pop-up, if desired.

Dashboard Tailored to me

Based on the integrated data and pregnancy lifecycle, the dashboard provides pregnant users with recommended actions for their health and the health of their baby, including necessary hospital visits and government support policies. Additionally, using this information, the system can automatically fill in the pregnancy diary, make hospital appointments, and offer psychological counseling through AI. If users agree, a unique ID is shared with retail or content providers, allowing them to receive personalized content and product recommendations as part of special offers.

NEXT Projects

B2B2C Chatbot Project
Finance, Retail chatbots