TOKENEED is a global blockchain tokenization platform connecting investors with vetted projects across its partner network. TOKENEED, a global tokenization ecosystem built on transparency, community empowerment, and regulatory compliance, needed to release product updates faster without sacrificing stability. ASCENDING designed and implemented a continuous deployment pipeline on Amazon EKS using AWS CodePipeline, replacing a slow, manual release process with a standardized, zero-downtime workflow.
BackgroundAn Agile Blockchain Tokenization Platform Running on Amazon EKS
TOKENEED is an Agile blockchain tokenization platform that runs its production workloads on Amazon EKS. TOKENEED bridges world-class blockchain projects with global investors through a complete tokenization system built for sustainable success across its partner and community network. As an Agile team, TOKENEED releases its product frequently, incorporating end-user feedback into each iteration to keep pace with a fast-growing market.
TOKENEED runs its application on Amazon EKS, containerizing and deploying it as Kubernetes pods, services, and deployments. Prior to this engagement, ASCENDING had already worked with TOKENEED on AWS architecture design and application development, building a foundation of trust in the team's professionalism, accountability, and delivery quality.
The ChallengeManual Docker Builds and Kubernetes Releases Without Rollback
The challenge is a manual, error-prone Kubernetes release process that could not keep pace with TOKENEED's growth or provide reliable rollback. As a fast-growing enterprise in the tokenization space, TOKENEED needed to update production frequently to keep up with market demand, new technical tooling, and evolving regulatory requirements. Its legacy release process could not keep pace.
- Every release required manually building a package from source code, creating a new Docker image, and delivering it to the Kubernetes cluster — a time-consuming and error-prone sequence.
- Without a systematic continuous deployment pipeline, releases were unstable and unstandardized across the team.
- Rolling back a previous version was difficult, adding risk to every production update.
Containerization and AWS Infrastructure Automation Experience
ASCENDING's team is highly experienced in containerization and AWS infrastructure automation, having delivered similar assignments for organizations across different scales. Because ASCENDING had already partnered with TOKENEED on AWS architecture design and application development, the team understood TOKENEED's environment and requirements coming into this engagement. That track record of professionalism, accountability, and productivity is why TOKENEED chose ASCENDING to lead this complex continuous deployment assignment.
The SolutionAn AWS CodePipeline Workflow from Source to EKS Deployment
The solution is an automated AWS CodePipeline workflow that moves every release from source control to a running Amazon EKS deployment. ASCENDING built a continuous deployment pipeline on AWS CodePipeline that carries every release from source code to a running EKS application, with monitoring across each stage.
- Source: Retrieve source code from GitHub and AWS CodeCommit.
- Test: Connect to a test database to run unit tests, either before or after the build stage depending on the release.
- Build: Build a .war file and a Docker image, storing them in Amazon S3 and Amazon ECR respectively, or passing them directly to the next stage.
- Deployment: Deploy the current version to the running application using a Rolling, Immutable, or Blue/Green strategy, ensuring zero downtime for users. AWS Lambda and AWS CodeBuild can both execute this deployment step, with the key action being an update to the running image version.
- Monitor: Track the status of each pipeline stage and the health of the application, with alerting throughout the entire continuous deployment process.
80% Faster Releases with Zero-Downtime Kubernetes Deployments
The outcome is measurably faster, more reliable releases built on Rolling and Blue/Green deployment patterns recommended by the AWS Well-Architected Framework. The new continuous deployment pipeline gave TOKENEED a faster, standardized, and more reliable way to ship product updates, translating directly into a quicker return on investment.
- Release time dropped by more than 80% compared to the legacy manual process, accelerating TOKENEED's time to ROI on new product features.
- Every product release now follows a consistent, standardized deployment behavior across the team.
- The architecture improved security, high availability, and stability for the production environment.
- Each deployment is tagged with a unique hash code, making it easy to trace back to and roll back to previous production versions — all with zero downtime for customers.
Built with AWS DevOps and Kubernetes Services
Frequently Asked Questions
How long does it take to stand up a CI/CD pipeline like this on Amazon EKS?
Timeline is typically a few weeks for a first production-ready pipeline, covering pipeline design, integrating source control, build, test, and deployment stages, and validating rollback behavior before cutting over from the manual release process.
How does the pipeline avoid downtime during a release?
Zero downtime comes from using Rolling, Immutable, and Blue/Green deployment strategies, orchestrated through AWS CodeBuild and AWS Lambda, so the running EKS application always has healthy pods serving traffic while a new version rolls out.
What happens if a deployment needs to be rolled back?
Rollback is fast because every release is tagged with a unique hash code, letting the team trace back to and redeploy a known-good version without manually reconstructing the previous build.
Does this pipeline fit regulated industries like blockchain and fintech?
Governance is built into the pipeline through standardized, auditable deployment steps and consistent monitoring across every stage, which supports the compliance and traceability expectations common in blockchain and financial technology environments.
