Transforming UrbanSim into an ML-Ready Platform
Challenge
UrbanSim is a platform that belongs to Paul Waddel, a leading expert in urban analytics and former professor at Berkeley University. The platform analyzes statistical data and predicts the optimal scenarios for better city planning.
Unable to scale the city planning platform further, Paul Waddel turned to Flyaps to rebuild it.
UrbanSim is a platform that belongs to Paul Waddel, a leading expert in urban analytics and former professor at Berkeley University. The platform analyzes statistical data and predicts the optimal scenarios for better city planning.
Unable to scale the city planning platform further, Paul Waddel turned to Flyaps to rebuild it.
Solution
By closely collaborating with UrbanSim, Flyaps delivered an ML-ready solution built using microservices architecture, cloud technology, and a modern tech stack.
- Helped to move from an outdated legacy solution to a modern scalable system
- Replaced monolith with flexible microservices architecture
- Achieved full DevOps automation of environments
- Prepared the infrastructure for MLOps, allowing machine learning adoption
- Reduced cloud costs by 50%
By closely collaborating with UrbanSim, Flyaps delivered an ML-ready solution built using microservices architecture, cloud technology, and a modern tech stack.
- Helped to move from an outdated legacy solution to a modern scalable system
- Replaced monolith with flexible microservices architecture
- Achieved full DevOps automation of environments
- Prepared the infrastructure for MLOps, allowing machine learning adoption
- Reduced cloud costs by 50%
Results
- No tech debt
- 40% lower maintenance costs
- 50% reduction in cloud expenses
The city planning platform is used across the US and internationally, with upcoming adoption in the US, Canada, and Australia. It can be scaled and customized for diverse locations through independent development streams.
- No tech debt
- 40% lower maintenance costs
- 50% reduction in cloud expenses
The city planning platform is used across the US and internationally, with upcoming adoption in the US, Canada, and Australia. It can be scaled and customized for diverse locations through independent development streams.