The evolution of fullstack engineers

The episode follows Mrina Sugosh’s path from coding on a TI-84 to IBM Cloud, then into developer relations. It uses that background to describe how fullstack work has shifted from a strict front end/back end split to developers who can move across client, server, data, and deployment tasks.

Reading notes

#

• Her path started in high school with a TI-84 calculator, then moved through C++, front end work, computer science studies, IBM Cloud, DigitalOcean, and now DevRel at CKEditor.
• Fullstack has become more common in startups and smaller teams where one developer may handle both client-side and server-side work, end-to-end setup, and tasks across the data pipeline.
• Shared languages like JavaScript make the move easier, especially with Node.js and ExpressJS reducing the gap between front end and back end work.
• Front end work has become heavier, but it still needs optimization through lazy loading, selective rendering, and query-based data access such as GraphQL.
• Front end security risks include cross-site scripting, clickjacking, and CDN tampering, while back end work centers on authentication, requests, and safer server-to-server flows.
• For front end growth, she points to vanilla JavaScript, web components, the DOM, accessibility, cross-browser compatibility, and tools like Vite.
• For back end growth, she points to Node.js, databases such as MongoDB, RESTful APIs, GraphQL, Docker, Kubernetes, and cloud services.
• She says back end work still includes server-side rendering, data processing, asynchronous reports, and algorithmic problem solving.
• Edge deployment can help with geographically closer delivery, but she thinks it fits some real-time and IoT uses better than heavy compute or large asynchronous workloads.
• She says fullstack roles can still fit large organizations, but the exact value depends on project structure, sprint needs, coordination costs, and how much depth the team needs in each area.
• She expects the role to keep expanding toward data, ML, AI, and DevOps-related work, but thinks SRE should remain its own team.
• She says AI can assist coding and improve productivity, but it still hallucinates and is not ready to replace fullstack engineers.
• Her closing point is that not everyone needs to become fullstack; people should keep working in the areas they are good at and interested in.