Problem Statement

The StreamDaily Team wanted a news application on which news from multiple news providers can be displayed based on users interests using a paid API. Also, the team wanted to showcase their own news on the website from their news media sources. The main aim of the team was to show all the news, both third-party and from own sources, in a short and meaningful way in the form of points. Also, as their news media is new, they also needed an application that allows it's editors to publish, edit and manage news from their media house. By understanding their requirements thoroughly we came up with a solution that is both robust and modern.

Technical Solution

So, giving emphasis on the term "Modern" and "Robust", we came up with a solution that uses multiple modern technologies, tools and libraries. We use the modern Typescript/JavaScript libraries and frameworks, AWS Serverless Stack with the integration of multiple LLMs into the solution. The News Website had to be quick, and the individual articles loading fast. For this we made use of modern server components that allows for fetching and rendering of the data on the servers thus only rendering HTML to the website users. Also, for the News Editor we made use of the minimalistic but customizable open-source WYSWYG library allowing for custom features and funcitonalities. Also, as the News App is relatively new, we made use of AWS Serverless Technology that charges the client for only the resources used. Thus no over-paying when the user base is small while also allowing for scalability as the user base grows in size over time. Also, with our expertise in Event-Driven-Services, we had use multiple Analytics functions to measure the Website Traffic, User Experience and Website usage.

Technologies Used

We have used modern technologies and tools for the project and also created IaC solution that can be easily replicated across multiple regions of the AWS cloud. The solution also includes a robust CI/CD pipeline that allows for quick changes and features addition. Below, we provide a list of technologies, programming languages and tools used.

• Programming Languages: Typescript, Python

• Frontend Libraries/Frameworks: React.js, Next.js, Jest, MUI, React-Testing-Library

• AWS Serverless: AWS Lambda, AWS EventBridge, AWS DynamoDB, AWS AppSync, JavaScript SDKs, Python SDks

• GenAI Tools: AWS Bedrock Serverless, LLama LLM, Anthropic Claude

• AWS Infrastructure: AWS Amplify, AWS CDK

• Miscellanous: Google Analytics, Web Vitals

Our Robust Methodology

We follow proven methodologies along with new techniques in our Application Development Process so as to meet right standards and practices. Our Development Team is lean and aligns well with the priciples of Agile Development Methodology. We allows quick sandbox deployments to test the solutions and gather analytics which helps decide wheter a solution/feature is feasible and useful. Below are few common practices we follow

• Requirements Enginnering: As Requirements change quickly, we develop prototype from the base minimum requirements, and make it customizable and pluggable

• Architectural Design: Choosing a proper and right Architecture is our priority as the entire project life-cycle depends on it. We invlove stake holders and technical team members. We choose the right architecture from the client requirements, technical knowledge, and long-term maintanence costs in mind.

• Systems Modelling & Documentation: Once the architecture is selected, we select the right tools and frameworks and model the system using diagrams and create an overview of how the individual tools will communicate with each other. We make sure our model is translated into code using the modern IaC methodology. We Document the system design and technical decisions simultaneously.

• Development & Testing: We develop individual modules and simultaneously conduct unit-test which helps on the long term during system integration & Testing and End-to-End Testing. We develop code with proper maintainable and readable code with proper and useful comments where necessary. We follow best practices keeping the eye on the open-source communities and do not stick to outdated methodologies and techniques.

• DevOps, Observability & Operations: Deploying modular changes following proper testing enables to add features progressively. We follow good DevOps practices and use tools that are industry tested. We use application logging services for early detection and resolution of issues that occur. Also, we use detectors when any resources behave abnormally to quickly fix it. We use LLMs to Q&A on the application logs, metrics and errors to find "why something failed and doesn't work as expected"

Testinomial

Working with the team has been a really fulfilling experience, they take complete ownership of the execution and go beyond their limits to suggest best solutions dynamically

—StreamDaily