Why Signeasy chose AWS Serverless to build their SaaS dashboard

Signeasy is a leading eSignature company that offers an easy-to-use, cross-platform and cloud-based eSignature and document transaction management software as a service (SaaS) solution for businesses. Over 43,000 companies worldwide use Signeasy to digitize and streamline business workflows. In this blog, you will learn why and how Signeasy used AWS Serverless to create a SaaS dashboard for their tenants.

Signeasy’s SaaS tenants asked for an easier way to get insights into tenant usage data on Signeasy’s eSignature platform. To address that, Signeasy built a self-service usage metrics dashboard for their SaaS tenant using AWS Serverless.

Usage reports

What was it like before the self-service dashboard experience? In the past, tenants requested Signeasy to share their usage metrics through support channels or emails. The Signeasy support team compiled the reports and then emailed the report back to the tenant to service the request. This was a repetitive manual task. It involved querying a database, fetching and collating the results into an Excel table to be emailed to the tenant. The turnaround time on these manual reports was eight hours.

The following table illustrates the report format (with example data) that the tenants received through email.

Figure 1. Archived usage reports

The design

Signeasy deliberated numerous aspects and arrived at the following design considerations:

  • Enhance tenant experience — Provide the reports to tenants on-demand, using a self-service mechanism.
  • Scalable aggregation queries — The reports ran aggregation queries on usage data within a time range on a relational database management system (RDBMS). Signeasy considered moving to a data store that has the scalability to store and run aggregation queries on millions of records.
  • Agility — Signeasy wanted to build the module in a time-bound manner and deliver it to tenants as quickly as possible.
  • Reduce infrastructure management — The load on the reports infrastructure that stores and processes data increases linearly in relation to the count of usage reports requested. This meant an increase in the undifferentiated heavy lifting of infrastructure management tasks such as capacity management and patching.

With the design considerations and constraints called out, Signeasy began to look for the suitable solution. Signeasy decided to build their usage reports on a serverless architecture. They chose AWS Serverless, because it offers scalable compute and database, application integration capabilities, automatic scaling, and a pay-for-use billing model. This reduces infrastructure management tasks such as capacity provisioning and patching. Refer to the following diagram to see how Signeasy augmented their existing SaaS with self-service usage reports.

Architecture of self-service usage reports

Figure 2. Architecture diagram depicting the data flow of the self-service usage reports

  1. Signeasy’s tenant users log in to the Signeasy portal to authenticate their tenant identity.
  2. The Signeasy portal uses a combination of tenant ID and user ID in JSON Web Tokens (JWT) to distinguish one tenant user from another when storing and processing documents.
  3. The documents are stored in Amazon Simple Storage Service (Amazon S3).
  4. The users’ actions are stored in the transactional database on Amazon Relational Database Service (Amazon RDS).
  5. The user actions are also written as messages into message queue on Amazon Simple Queue Service (Amazon SQS). Signeasy used the queue to loosely couple their existing microservices on Amazon Elastic Kubernetes Service (Amazon EKS) with the new serverless part of the stack.
  6. This allows Signeasy to asynchronously process the messages in Amazon SQS with minimal changes to the existing microservices on EKS.
  7. The messages are processed by a report writer service (Python script) on AWS Lambda and written to the reports database on Amazon Timestream. The reports database on Timestream stores metadata attributes such as user ID and signature document ID, signature document sent, signature request received, document signed, and signature request cancelled or declined, and timestamp of the data point. To view usage reports, the tenant administrators navigate to the Reports section of the Signeasy portal and select Usage Reports.
  8. The usage reports request from the (tenant) Web Client on the browser is an API call to Amazon API Gateway.
  9. API Gateway works as a front door for the backend reports service running on a separate Lambda function.
  10. The reports service on Lambda uses the user ID from login details to query the Amazon Timestream database to generate the report and send it back to the web client through the API Gateway. The report is immediately available for the administrator to view, which is a huge improvement from having to wait for eight hours before this self-service feature was made available to their SaaS tenants.

Following is a mock-up of the Usage Reports dashboard:

Figure 3. A mock-up of the Usage Reports page of the Signeasy portal

So, how did AWS Serverless help Signeasy?

Amazon SQS persists messages up to 14 days, and enables retry functionality for message processed in Lambda. Lambda is an event-driven serverless compute service that manages deployment and runs code, with logging and monitoring through Amazon CloudWatch. The integration of API Gateway with Lambda helped Signeasy easily deploy and manage the backend processing logic for the reports service. As usage of the reports grew, Timestream continued to scale, without the need to re-architect their application. Signeasy continued to use SQL to query data within the reports database on Timestream in a cost optimized manner.

Signeasy used AWS Serverless for its functionality without the undifferentiated heavy lifting of infrastructure management tasks such as capacity provisioning and patching. Signeasy’s support team is now more focused on higher-level organizational needs such as customer engagements, quarterly business reviews, and signature and payment related issues instead of managing infrastructure.

Conclusion

  • Going from eight hours to on-demand self-service (0 hours) response time for usage reports is a huge improvement in their SaaS tenant experience.
  • The AWS Serverless services scale out and in to meet customer needs. Signeasy pays only for what they use, and they don’t run compute infrastructure 24/7 in anticipation of requests throughout the day.
  • Signeasy’s support and customer success teams have repurposed their time toward higher value customer engagements vs. capacity, or patch management.
  • Development time for the Usage Reports dashboard was two weeks.

Further reading

https://aws.amazon.com/blogs/architecture/why-signeasy-chose-aws-serverless-to-build-their-saas-dashboard/

Leave a Reply