Project Overview
Walla Software aimed to enhance its fitness studio management platform with advanced predictive capabilities to reduce client churn. Facing significant data volumes from studio visits, bookings, and membership details, the company needed a solution that could clean and prepare data for machine learning while providing real-time predictions. Avahi developed an AI-driven pipeline on AWS that ingests raw client information into Amazon RDS, processes it through AWS Lambda, stores cleansed data in Amazon S3, and trains predictive models in Amazon SageMaker. This streamlined architecture delivers actionable insights, helping studio owners proactively engage clients and boost retention rates.
About the Customer
Walla Software offers a modern management solution for boutique fitness studios. Its platform combines scheduling, payment processing, and client relationship tools to streamline daily operations. With rapid growth and an expanding customer base, Walla Software needed a scalable way to leverage client data for predictive analytics.
The Problem
Manual processes and siloed data limited Walla Software’s ability to identify at-risk clients before they churned. The company stored critical information—such as visits, bookings, and membership plans—in disparate systems, making it challenging to build accurate, timely churn models. As membership numbers continued to grow, so did the complexity and urgency of implementing a robust, automated predictive solution.
Why AWS
AWS offered an integrated suite of services that could quickly scale to handle Walla Software’s data and analytics needs. Amazon RDS simplified data ingestion and management, while Amazon S3 and AWS Lambda streamlined data processing. AWS SageMaker provided a fully managed environment to train, validate, and deploy machine learning models. This comprehensive toolset minimized infrastructure overhead and sped time-to-market.
Why Walla Chose Avahi
Walla Software engaged Avahi for its deep expertise in cloud-native development and AI-driven solutions. Avahi’s proven track record with AWS services and agile methodology aligned well with Walla Software’s need for rapid, high-impact deliverables. By collaborating closely with the Walla team, Avahi ensured every step of the engagement was tailored to Walla Software’s specific operational requirements and data security standards.
Solution
Avahi built a data pipeline leveraging Amazon RDS to store and organize client information. AWS Lambda functions then cleaned and transformed this data before uploading it to Amazon S3. Amazon SageMaker used the prepared datasets to train and refine the churn prediction model, producing actionable insights for studio owners. For real-time predictions, the model was exposed through Amazon API Gateway, with AWS Lambda handling incoming requests. Finally, Amazon CloudWatch provided detailed metrics and alerts, enabling continuous monitoring and performance optimization.
Key Deliverables
- Centralized data storage in Amazon RDS for client and membership records
- Scalable model training in Amazon SageMaker
- Real-time prediction endpoints via Amazon API Gateway and AWS Lambda
- Knowledge transfer and documentation for ongoing improvements
- Automated data cleaning using AWS Lambda
- Secure data storage in Amazon S3 with clear retention policies
- Continuous monitoring and alerts in Amazon CloudWatch
Project Impact
Walla Software’s new predictive analytics pipeline empowers fitness studio owners to proactively address client churn. Automated data processing and machine learning capabilities now offer near real-time insights, leading to more personalized outreach and higher client satisfaction. The unified AWS environment reduces operational complexity, enabling Walla Software to focus on innovation rather than infrastructure.
