Delivering Providers Near-Instant Patient Insights Using Amazon Aurora with Bamboo Health

Bamboo Health has terabytes of patient data that it securely manages using Aurora, which it selected for its performance and scalability. The company’s solutions process around 5 million patient transactions daily—averaging over 3,400 transactions per second across databases for each state. There are stringent compliance regulations around the security of the data, and the application needs to be highly available so that care providers and pharmacists have access to this critical patient data. “Aurora gives us massive reliability and scalability,” says Owen Marshall, director of product platform at Bamboo Health. “It also has baked-in answers to many compliance-related aspects.” Because Aurora is compliant with major regulations, Bamboo Health reduces the time it takes to prove compliance to auditors for HIPAA regulations and to maintain its HITRUST CSF certification.

Because the solution is so critical to providers, updating its database was difficult for Bamboo Health. Major upgrades to the latest version of PostgreSQL would take months of planning and hours of downtime. “There was just no real way for us to keep moving forward with major version upgrades without massive client pain,” says Marshall.

The solution was blue/green deployments, a feature of Aurora that is used to make changes to the database in the staging environment without affecting the production environment. Bamboo Health worked alongside AWS Enterprise Support—optimized cloud services for enterprises—to create a proof of concept for the feature, which it ultimately adopted. “This blue/green model is the only way that we’re doing database maintenance now,” says Marshall. “It makes life so much easier for us.”

To use blue/green deployments, Bamboo Health clones its Aurora storage volume, which becomes the “green” version of its database. It upgrades the green version, then switches production from using the previous “blue” version to using the “green,” upgraded version. During implementation, Bamboo Health first rolled out blue/green deployments in lower-traffic regions, then scaled them out for its entire database environment. Bamboo Health worked alongside AWS Enterprise Support to keep costs low for the proof-of-concept phase, which the two planned and completed together in 2 months.

Blue/green deployments reduce the time needed for updates from hours of planned downtime to mere minutes. Thus, Bamboo Health can upgrade its databases without customers going several hours without access to critical patient data. “We don’t even have to have an outage window; we can do database maintenance during business hours,” says Marshall. “We could never do anything like this before—not even close. Now, updates are a routine part of our standard operating procedure.”

Now that upgrades are quick and simple, engineers can redirect energy from planning outages to optimizing performance or pursuing new use cases. It also means that Bamboo Health is up-to-date with the latest versions of PostgreSQL and Aurora, which empowers it to use the newest features, such as by modernizing its extract, transform, load (ETL) process. “Being on modern versions of the databases frees up engineers to use new functionality that AWS is rolling out,” says Marshall. “It also keeps our systems at an appropriate security posture, which is huge because we have so much protected healthcare data in these systems and it’s our priority to keep this information secure.”

One key feature that Bamboo Health can access now that it is using the latest version of Aurora is Amazon Aurora zero-ETL Integration with Amazon Redshift, which facilitates near real-time analytics on petabytes of transactional data. Within seconds of transactional data being written into Aurora, the zero-ETL feature seamlessly makes the data available in Amazon Redshift, which powers modern data analytics at scale. The feature removes the need to create and manage complex data pipelines to perform ETL operations, which Bamboo Health estimates would have taken 6 months to build. Zero-ETL is the backbone of the ETL process in a new solution that Bamboo Health will roll out in 2025. “Zero-ETL functionality is exciting for us because we’ve got a small number of engineers, and with this feature, they can actively make valuable data solutions rather than moving data from point to point,” says Marshall. “It’s one more thing that is already built that we don’t have to build ourselves.”

Using Aurora, Bamboo Health manages over 200 databases with 10 cloud engineers. “Aurora has database replication across multiple Availability Zones, which slots right into our compliance processes and designs,” says Marshall. “We don’t have to worry about scaling or replicating databases; we just know it works. That is how Aurora supports Bamboo.”

Bamboo Health is upgrading to the latest version of PostgreSQL using blue/green deployments and using the opportunity to migrate to AWS Graviton processors, a family of processors designed to deliver the best price performance for cloud workloads. “AWS Graviton is cost-effective, and the performance is perfectly in line with what we need,” says Marshall. The company saved 30 percent on database costs using Aurora I/O-Optimized cluster configurations for improved price performance and predictability for its I/O-intensive applications.

“If you’re investing in managed services on AWS, lean in and use them to the full extent of the capabilities,” says Marshall. “Problems often disappear from my road map because AWS is already implementing new functionality that addresses them. By staying current with the latest versions, we’re unlocking so much value from our technology and improving what we can deliver to our clients.”