Katy Reed

Eddy

Eddy is an AI learning companion built around structured conversational flows and activity-based learning. It is currently in internal rollout after a successful Open University pilot with strong session time and completion rates.

YEAR

2024 - 2025 Ongoing

Role

Product Design Lead, Principal ownership of conversational UX

TEAM

Scope

Learning flows, user journeys, personalization

IMPACT

Higher engagement and reduced creator effort

Where the learner begins the interaction with Eddy.

Business Need

The company required an AI-driven model that improved engagement and reduced creator workload. Previous dialogue based solutions were manual, inconsistent, and did not scale.

Team &
Collaboration

Worked with product, engineering, content teams, guided 4 product designers across interaction, animation, and character work. I set the UX direction, defined priorities, and guided the core design work end-to-end.
I also kept the senior designer’s workstream aligned with the overall product direction.

Lesson flow at a glance.

Responsibilities

Defined conversational architecture and core UX rules.
Designed learning progression, feedback timing, and interaction patterns.
Built fallback logic and personalization rules.
Coordinated delivery across engineering, product, and content.

Adaptive activities that match any topic or level.

Key decisions

Defined the structure of the learning experience across chat, activities, and feedback.
Shaped character behavior, activity patterns, and UX responses for AI ambiguity.

Learner view of the plan and progress path.

Outcomes

18 minutes average session time in Open University pilot with high completion rates.
Reduced creator effort through structured activity templates.
Established a reusable conversational model now in internal rollout.

From early sketches to final characters.