I question familiar ways of working, then solve problems with data and AI.

AI initiatives were moving independently across the organisation, without shared direction or priorities. I brought products, people and work processes into one strategy, proposed an execution roadmap and dedicated team, and helped lead to the creation of a company-wide AX organisation.

Strategy

Company-wide AX criteria

Execution model

Diagnostic · roadmap · org design

Organisational change

Dedicated AX team established

• Established company-wide AX criteria spanning products, people, work processes and shared foundations
• Designed a diagnostic model and execution roadmap that separated the current stage from the next target
• A dedicated AX organisation was established after the executive presentation

Employees' daily work, team performance and management decisions were disconnected by separate reporting processes. I analysed work by role and the plan–execute–review cycle, identified the bottlenecks, and designed core scenarios and a pilot in which AI connects planning, execution and leaders' decisions.

Scope analysed

Company-wide work processes

Core scenarios

8

Pilot priority

4 workflows

• Redesigned work processes so goals, execution and outcomes connect from employees to executives
• Defined 8 core scenarios and service features that divide responsibilities between people and AI
• Proposed 4 priority workflows and functional validation criteria for the pilot

There was no evidence for deciding on an AI rollout for 450 employees. Starting with actual usage data from 89 engineers, I estimated productivity impact by role and usage level, compared the costs and benefits of three rollout options, and recommended one.

Survey evidence

89 engineers

Rollout scope assessed

450 employees

Estimated ROI

319%

• Measured actual usage levels and perceived productivity through an 89-person survey
• Estimated cost and productivity impact across 450 employees by role and usage level
• Recommended an option with 319% ROI and a 2.9-month payback

More than ₩500M in ad spend had to be allocated more precisely while increasing repeat purchases. I connected marketing mix modelling (MMM), ad-performance decline prediction and purchase-history-based customer segmentation (RFM) to decisions, improving customer acquisition cost (CPA) efficiency by 11% and repeat-purchase conversion by 20%.

Repeat purchase

+20%

CPA efficiency

+11%

Ad spend managed

₩500M+

• Improved customer acquisition cost (CPA) efficiency by 11% with marketing mix modelling
• Used a model predicting ad-performance decline to decide when to continue or replace creatives
• Increased repeat-purchase conversion by 20% through purchase-history-based customer segmentation

Data and definitions from sales through settlement were scattered, so they had to be assembled again for every meeting. I aligned the workflows and KPIs first, then automated collection, aggregation and visualisation so the evidence was ready when the meeting began.

Aggregation lead time

2 hrs → 15 min

Aggregation time

87% reduction

Daily meeting

30 min shorter

• Defined company-wide workflows and KPIs through staff interviews, then automated collection, aggregation and visualisation
• Aggregation lead time 2 hrs → 15 min (87% reduction)
• Delivered decision data before meetings, cutting daily meeting time by 30 minutes