Personalized daily AI content inside Amazon

Accelerate is a leadership development program for Amazon L4 and L5 leaders, built on well-defined competencies and delivered in a hybrid format (in person plus digital). Despite how robust the program was, there was a clear challenge: keeping development active day to day, outside the in-person sessions.

Problem:

• Learning was concentrated in the live sessions
• Complementary content was generic and rarely used
• There was no continuous reinforcement based on individual needs
• Leaders had little time and little motivation to consume long materials

Solution:

I created Readings of the Day, a tool that generates 3 personalized readings every day using AI.

• Each reading is based on the user's weaknesses, career goals and learning path
• The content takes into account the company's context and how it defines leadership
• The readings are short, practical and aimed at immediate use
• Users consume them in the flow of their day, without friction The goal was not to create more content. It was to create the right content at the right time.

Technologies used:

• AWS Bedrock with Claude (to create the content)
• AWS Lambda Article reading screen

Reading of the day: new content every day, hyper-customized.

The core problem was not a lack of content. It was a lack of context and continuity. The main decision was to connect three elements that are normally kept apart:

• Competency assessment
• Individual goals
• Educational content Instead of a static library, I designed a system where content is born from the user's context. This made it possible to:
• Cut the waste of irrelevant content
• Ensure alignment with the company culture
• Create an experience that evolves along with the user In the long run, this approach turns content into an adaptive system, not a repository.

User development journey

The experience was designed to fit the real routine of an operations leader.

• Content appears automatically, with no need to search
• Each reading takes a few minutes
• The text speaks directly to the user's context
• There is a simple cycle: read, reflect, apply I avoided any unnecessary complexity. The priority was to make consumption natural and recurring, not "too educational". The day's available readings

Features

The solution evolved with a focus on relevance and consistency:

• Daily generation of 3 personalized readings
• Content adapted to the user's weakest competencies
• Integration with the active learning path
• Contextualization with the company's culture and expectations
• A structure with practical tips and guided reflection During development, the main improvement was moving from generic content to content that "speaks to the user".

Results

• More frequent contact with educational content
• A stronger perception of relevance among users
• Less dependence on in-person sessions
• A light habit of continuous development Key lesson: content doesn't engage because of its quality alone. It engages through context, timing and immediate usefulness.