Immunis.AI required a cloud services provider to seamlessly architect and implement its early-stage genomic data storage and analysis solution on AWS. This partner needed to bring the expertise to manage DevOps initiatives and optimize cloud costs, while also streamlining Immunis.AI’s data and analytics ingestion to enable viability as a commercialized platform.
Mission met these goals and is helping Immunis.AI to begin to commercialize its first of several clinical tests by introducing managed DevOps automation to ingest and wield cancer trial data with efficiency. Immunis.AI then leveraged Mission Cloud Foundation to implement AWS best practices and cost optimization; Mission was able to achieve significant savings and operating margins well beyond what Immunis.AI was expecting .
While it is still evolving, Mission provided Immunis.AI with a tested and proven blueprint for a viable research oriented genomic platform, all backed by AWS and ready to scale quickly and economically. The company is now expanding its Mission-built AWS infrastructure and replicating the initial success with plans for accelerated growth throughout 2021 and beyond.
“We would have had to hire three people to get what we’re getting from Mission, and even then wouldn’t have reached the level of experience that Mission brings. I look at Mission as an extension of our team. It’s a true partnership. Mission’s experience and expertise is extremely valuable to us.”
Senior VP Corporate Development
A biotech startup developing a blood-based diagnostic biomarker platform, Immunis.AI realized early on that it needed an AWS services provider to capably bring its solution into the cloud, and to build out infrastructure that could scale and be easily accessible to Immunis.AI’s global team. Initially starting from a file-based system with a mature application yet to come – and with minimal cloud literacy among the internal team – Immunis.AI required outside expertise to harness the power of the cloud.
Immunis.AI recognized that success would hinge on leveraging AWS to achieve rapid and efficient data ingestion, as well as an ability to leverage DevOps automation and cloud optimization to minimize costs wherever possible. The company also wanted to move its data to Amazon S3, and needed help navigating heavy data management workloads.
In early 2020, Immunis.AI began working with Mission, with the ultimate goal of running a production lab by the end of 2021. The stakes were high: the initial clinical trials would serve as a proving ground for Immunis.AI’s platform as an eventual commercial offering. From overcoming the weeks-long manual process of ingesting genomic data to Amazon S3, to controlling spikes in AWS costs due to usage anomalies across Immunis.AI’s unique implementation, Mission would need to manage a series of steps to ensure the startup would successfully hit its performance and timeline goals.
Immunis.AI was advised to explore AWS by the CISO of another biotechnology company, who knew the Immunis.AI leadership team.
“Given our industry, and other experiences I’ve heard in the world of biotech and diagnostics, AWS was the clear choice,” said Erickson. “By beginning to implement AWS so early as a company, we understood that we were, perhaps, buying a Cadillac before we could drive it. But we also knew that in our industry, you can’t start early enough. Companies in this space have gotten severely tripped up by not being prepared with the platform, architecture, and systems in place to manage growth. We knew our early investment in AWS would pay off, because of the speed at which we’re now able to add sophisticated modules to our platform.”
After viewing early plans for Immunis.AI’s solution, their advisor also recognized the company’s need for an expert cloud services provider that could better ensure the success of Immunis.AI’s blueprint initiative. The advisor, who had also worked with Mission at his biotech company, recommended Mission for its deep AWS acumen and hands-on support helping customers achieve their unique cloud goals.
“Part of our strategy was building up our own education,” said Erickson. “Instead of buying hardware, which would easily cost us $200,000 in equipment plus data center costs, we’ve learned and grown with AWS. There’re so many nuances in how you use the platform that I don’t think you can learn unless you’re doing it and, frankly, making mistakes. We’ve been glad to have Mission to guide us.”
“We began as a startup with a bunch of laptops on desks, and a blank slate when it came to infrastructure. What I quickly learned about Mission was its ability to bring us into the 21st century and get our operations up in the cloud, and fast. We’ve got employees everywhere, from Southern California to Belgium. Having one central IT hub in the cloud made a lot of sense for us.”
Senior VP Corporate Development
With the goal of launching Immunis.AI’s first clinical test by 2021, Mission set out to implement an AWS architecture built to deliver the functionality, scalability, compliance, and cost efficiency the solution required.
Mission built an AWS infrastructure for Immunis.AI utilizing Amazon S3 alongside a backend data pipeline, using Amazon EC2 and EBS infrastructure, that had been created by a third-party provider hired by Immunis.AI. Mission then focused on security, tightening the integrity of Immunis.AI’s public containers on Amazon EKS, and ensuring the biotech’s data arrives free of protected health information (PHI). Mission also continues to perform a full security scan across Immunis.AI’s platform each month, and corrects any issues. To cope with heavy data management workloads, Immunis.AI asked Mission about their current use of Microsoft SharePoint as a solution. Mission redirected the company to Amazon WorkDocs, and implemented the solution to enable efficient and secure storage and management of data content.
Immunis.AI and its sequencing providers use Illumina BaseSpace as a repository for generated data, which requires uploading data to BaseSpace and then ingesting this data from BaseSpace to Amazon S3. However, this was initially an entirely manual process. Processing data from BaseSpace to Amazon S3 would take nearly two weeks to complete when performed manually. To address this, Mission provided managed DevOps expertise that could take advantage of automation. Ultimately, the high-volume data ingestion process that took weeks is now completed in just hours.
In June 2020, as Immunis.AI experienced rising across-the-board AWS costs, Mission introduced the company to Mission Cloud Foundation. With this service, Mission Cloud Analysts performed a Cloud Optimization Review. As a result of this review, Mission set up Immunis.AI to benefit from AWS Savings Plans and Mission’s Instant Discount Program, which provided an immediate 5% savings. These optimization efforts brought Immunis.AI’s AWS costs in-line with the margins necessary for a commercialized product, and then some.
In late 2020, an issue occurred in which misutilization of Amazon EBS volumes resulted in a rapid increase in Amazon EBS and EC2 expenses. Mission traced the cause of the issue to the third-party provider operating a backend data pipeline. The provider was creating huge Amazon EBS volumes that stored entire data batches from Amazon S3 and misidentified how much space was truly needed, resulting in expensive overprovisioning. Mission proactively organized a meeting with Immunis.AI and the third-party provider to discover the root of these issues and educate the provider on best practices for utilizing the Amazon EBS volumes. It addressed the issue by introducing AWS Batch, a batch system able to read data from Amazon S3 and utilize spot instances to store and compute data far more efficiently. Mission Cloud Analysts also implemented Amazon S3 Intelligent-Tiering to auto-adjust cloud storage costs based on data retrieval patterns. Thanks to these improvements delivered through the Mission Cloud Foundation Service, Immunis.AI saw a significant monthly decrease in Amazon EBS and Amazon EC2 costs.
“When I saw our AWS bill in November, I thought I was going to have a heart attack. Within a month, Mission’s team helped us reduce that bill by about 65%. And performance stayed the same. It made a huge difference for us.”
Senior VP Corporate Development
With its streamlined and cost-optimized Mission-built AWS infrastructure in place, Immunis.AI completed its first clinical trial “in the cloud” at the end of 2020. Backed by DevOps automation and Mission Cloud Foundation, the trial was a complete success as a proof-of-concept, demonstrating even more operational efficiency and margin on business costs than Immunis.AI required.
Immunis.AI now has the blueprint for its genomic platform that is tested, proven, viable, and prepared to scale. With Mission’s Managed DevOps service and ongoing optimization, Immunis.AI is confident that the cost controls, security, performance, and AWS best practices it relies on will continue to provide an efficient and effective infrastructure as it expands its operations exponentially.
Immunis.AI plans to replicate its successful blueprint across many clinical studies throughout 2021 and beyond, as it develops and commercializes its products. Mission is continuing to support and refine Immunis.AI’s AWS infrastructure, preparing the company for full-scale commercial growth without needing to navigate cloud complexities in-house. For example, Mission is introducing Apache NiFi as an ingestion pipeline able to manipulate incoming data and ingest entire projects at once, which will yield tremendous gains in speed and performance. Mission will also continue to help Immunis.AI with data modernization, including data lake and analytics initiatives on AWS.
Immunis.AI is the developer of a non-invasive blood-based immunogenomic platform licensed from Harvard University. The biotech startup performs detailed genomic analysis of patient blood samples to understand diseases that individuals may be harboring (and the body’s immune response to that disease). From that analysis, Immunis.AI tells physicians whether a patient has a disease, what stage it is in, and whether it is benign or requires definitive treatment. Immunis.AI’s platform is initially focused on oncology, with future plans to broaden its focus to other diseases and treatments.