13 July 2026 · Open Engineering
From Snake Game to AI-native Front-end Architecture
A personal story about curiosity, abstraction, frontend foundations, AI/ML, and why the interface remains the most durable part of software.
Hello! I am Jangya. My full name is Jangya Narayana Satapathy.
It started unconsciously. I was fourteen when I got my first digital phone. There was a snake game on it, and I could not stop wondering how that pixel moved on a tiny screen, who wrote the rules, and whether I could too.
That itch never left. In college, while my classmates attended lectures, I spent nights in front of the computer teaching myself HTML and JavaScript, building terrible websites I was inexplicably proud of. Those sleepless nights, fueled by curiosity and instant noodles, became the foundation of everything I do today.
A decade later, that curiosity has carried me through the full evolution of modern front-end.
My first serious professional chapter was at Adept Concept Private Limited, a startup where I worked with jQuery, PHP, HTML, CSS, and JavaScript. It was not glamorous, but it was foundational. I was close to the browser, close to the DOM, close to the request-response cycle, and close to every mistake I made.
That early startup work taught me something I still carry: abstractions are useful only when you understand what they are hiding.
Looking back, the real story was not jQuery to Angular to React. It was abstraction. jQuery abstracted browser inconsistencies and DOM manipulation. Angular abstracted application structure. SAP UI5 abstracted enterprise UI patterns. React abstracted rendering into components and state. Every new frontend library is an argument about which complexity should be hidden, which complexity should be exposed, and who should carry the burden: the framework, the developer, or the user.
At Cognizant, I worked as a consultant for Apple, and that experience changed my sense of what good UI means. I got deep exposure to accessibility, design discipline, and performance: three things that sound like checkboxes until you see how much they shape trust. I also explored enterprise monoliths and Angular-era application structure, where the challenge was not just making screens work, but making them work consistently inside large systems.
At SAP Labs, I went deeper into SAP UI5, Angular, and MEAN-stack ideas across Node.js and MongoDB. I also got hands-on exposure to Docker, Kubernetes, and cloud-oriented delivery. That is where front-end stopped feeling like a layer at the edge of the system and started feeling like part of the system itself.
The moment we moved complex business logic to the browser, front-end stopped being a “layer” and became the application itself, and demand for people who could architect that complexity exploded.
At Oracle, I worked on enterprise-grade UI systems for Oracle Cloud, using framework-driven patterns built for large product surfaces. It deepened my understanding of distributed architecture and the boundaries where front-end meets platform services.
Understanding how your UI communicates with services, how containers package and isolate build artifacts, and how the cloud amplifies both your reach and your failure surface: these are not purely backend concerns anymore. They are frontend concerns too.
Now at Edelman Financial Engines, as a Staff Software Engineer, I lead front-end architecture and GenAI integration at scale. The stack has evolved to React, Zustand, and AI-powered advisory platforms, but the underlying question remains the same:
How do you build a UI that is fast, accessible, and honest about what it is doing?
Here is what I believe: UI is the last pillar that holds any application upright.
Backend services can be swapped. Databases can be migrated. AI models will be iterated on monthly. But the interface between human intention and digital capability, that layer is permanent. No amount of AI generation changes the fact that someone has to decide what the user sees first, how errors are communicated, when to delegate to an AI, and when to keep a human in the loop.
Those are architectural decisions. They require judgment, taste, and the lived experience of knowing what breaks in production.
That conviction is what drove me to pursue an M.Tech in AI/ML at BITS Pilani. Not to chase the hype, but to understand the technology from the inside out, so I can write about where AI and front-end architecture actually intersect, not where the hype says they do.
That foundation matters even more in the AI era. AI can generate code, but it cannot save you from a weak mental model. If you do not understand how the browser works, how state flows, why accessibility fails, or where performance breaks, you cannot debug what AI gives you.
My belief is simple: if you cannot build it, you cannot truly debug it.
The most interesting work in our field is not about what AI can generate. It is about what it cannot replace.
I live in Bengaluru with my wife and son.
When I am not architecting front-end systems or studying neural networks, you will find me at the gym, meditating, or chasing the next adventure, anything that reminds me that the physical world is still the most compelling interface there is.