About Customer
Increase the appeal of text blogs by complementary short format videos, grab customer attention with concise marketing videos, improve SEO ranking generate product/services buzz with Pictory. Over 50k users across the world, self-employed to enterprises, amateurs to social media influencers, have already experienced the power of AI powered pictory app.
Pictory provides an online application for creating short, branded videos for social media marketing. It does so by converting text to video as well as converting long format videos into engaging short videos. It’s advanced Artificial Intelligence (AI) automatically extracts ‘golden nuggets’ hidden deep within long format videos such as Zoom, Teams & Webinar recordings. Text blogs can be quickly converted into videos by adding stock footage, background music, captions in a matter of minutes.
Challenges
The application is developed using microservices architecture and deployed on AWS. It uses many PaaS and AI services like ECS and ECS Fargate, Cognito, Amazon Lex, Polly, Comprehend, Media Converter, etc.
- With significant growth in application usage, the customer wanted to ensure the application and the underlying platform are secured
- Cloud services are optimally configured to keep monthly AWS cloud spending optimized
- Improve time to market by launching new services and environments very quickly for testing new features and bug fixes
Approach and Solution
Codincity approach to address the challenges involved a holistic cloud platform and application assessment, resulting into a set up recommendations across security, cost and operational excellence pillars. The solution(s) involved implementing recommendations based on impact, cost-benefit analysis considering over all application design, technology constraints, and business objectives.
Architecture Best Practices
Application design followed industry best practices and developed using cloud native, serverless and PaaS services. However, cloud services for auditing, DDoS protection, access control using security services such as WAF, GuardDuty and appropriate IAM permissions were not fully aligned to best practices. Codincity identified 20+ such deviations from “well architected”, quantified impact and executed remediations to make the platform secure and internet ready.
Cost Optimization
Customer application was deployed using PaaS and serverless architecture. It was assumed that using serverless automatically improves costs. However, for a larger number of invoking of lambda functions, ECS tasks, the ECS Fargate is far from optimal when it comes to overall compute costs. The codincity team suggested using ECS EC2 for services, with compute savings plan, resulting in approx. 10% cost reduction.
IaC for Environment Provisioning
The application was developed and deployed using code pipeline. However, the AWS services used by the application were deployed and managed manually through console. Codincity team reverse engineered the services to come up with terraform code for the provisioning of services. Customer now has the ability to create new on-demand environments for feature testing/bug fixes, ensuring high application availability.
