
Kiu and Sessionize: Transforming Event Management in AWS User Groups
In this article, you will learn a bit about the history of Kiu and how it integrated with the Sessionize API
Published Feb 4, 2025
The story of Kiu's beginnings goes back to 2024, when its creator Hazel Saenz (an AWS Hero from Guatemala) conceived the idea of building a virtual agent that could initially be used by AWS Latin American communities and later expand worldwide.
From this initiative, Hazel began developing a virtual assistant project with WhatsApp integration and using Amazon Bedrock as the Generative AI engine, initially with information from the Guatemala User Group.
During August and September 2024, AWS Community Day events were taking place in various Latin American cities, and it was at this point that Hazel made the chatbot available for use at these events and to include more communities. Hazel asked the AWS Women Colombia User Group through their leader Ana Milena Barragán if they were interested in the project, to which she responded yes and referred me to join the team. That's where I met the virtual assistant Kiu (which at that time didn't have that name yet and was in its second pilot version).
I was very interested in being part of this idea and joining the team behind Kiu. Although I had previous experience developing chatbots with Amazon Lex, I hadn't yet had the opportunity to use Amazon Bedrock, so I was very excited to learn how to use it. I found Hazel's proposal very well thought out - centralizing information about speakers and their respective talk schedules for each event in a single agent, plus being able to show additional relevant data as needed
One of the most relevant tasks I was in charge of was integrating the sessions (or talks) from the Sessionize API into the agent to return information via WhatsApp.
To achieve this, I first had to understand the architecture of what was in place at that time. Kiu is built using IaC (Infrastructure as Code) that defines a complete AWS architecture for a virtual assistant integrated with WhatsApp. The system uses multiple AWS services, including: Secrets Manager for credentials, DynamoDB for storage, Lambda for processing, and Amazon Bedrock for Artificial Intelligence.
The architecture is modularized into different components that handle AI-based chat responses using Claude models. The entire system is automatically deployed using CDK, converting Python code into AWS CloudFormation templates.
Regarding session integration, I created a new Lambda function, structured in three main components that work together to provide accurate and organized data about sessions and speakers:
- The first component is the main handler (lambda_handler), which acts as the system's entry point. This element receives the request containing the Sessionize API GET sessions endpoint URL and coordinates the complete data retrieval and transformation process.
- For data retrieval, I implemented (get_sessionize_data) using urllib3 (Python HTTP request library). This function handles communication with the Sessionize API and manages responses, including error handling to ensure system robustness.
- The last function handles information processing (process_session_data), transforming raw data into a structured format:
- Talk Title
- Talk Description
- Speakers' First and Last Names
- Start and End Times
- Room Location
GitHub Repository:
Once the session integration module was completed, our dear friend Kiu was ready to fly from Guatemala to Colombia to help at the AWS Community Day Colombia https://awscommunitydaycolombia.com/. This event took place on August 24, 2024, and was successfully used by event participants to help them find their sessions. It could also answer more complex questions, such as when asked about sessions of interest based on a particular area.
Interesting data from AWS Community Day Colombia:
- Number of attendees: +800
- Number of talks: 40
- Workshops: 3
- Amazonian speakers: 10
(Data provided by Ana Milena Barragán)
Note: The text mentions screenshots of conversations with Kiu here



With this session integration, it was used for other Community Day, the next days, in Guatemala, Mexico, Chile, and other Latin American countries.
What's expected for Kiu's future? Kiu will have more trips around the world, our team is working hard to grow and implement new functionalities that will be very useful for the entire AWS community and can be used in future events.
It's also very interested in continuing to learn new languages, where perhaps its good friend Duo (Duolingo's mascot) will help with this task, such as English, French, Italian, German, Chinese, Japanese, and many more.
You can try Kiu here and interact with it https://www.kiu-ai.com/
Thank you for reading the article. If you're interested in knowing more news about Kiu, don't forget to follow our team, you can also share in the comments what new features you'd like to see in Kiu and which distant lands you’d like Kiu to explore. All comments will be taken into consideration.