Codegen in a Server Driven UI

Mar 21, 2024

Codegen in a Server Driven UI

This article breaks down the talk by Ayush Singh, Frontend Enigneer II at Disney+Hotstar, in our recent React Native Meetup at GeekyAnts Bengaluru.

Author

Ahona Das
Ahona DasSenior Technical Content Writer

In this discussion, the focus is on a different aspect of engineering workflows, namely, the potential development of a code generation platform within a server-driven UI. Ayush Singh delves into the intricacies of this concept and explores how it could revolutionize the way UI development is approached.


Understanding the Problem Statement

The talk begins by painting a picture of a scenario where product designers could effortlessly translate their designs into actual code. This seamless transition from design to implementation could significantly streamline the development process. However, achieving this level of integration between design and engineering requires careful consideration and innovative solutions.


The Need for Unification

In large organizations, product designers and engineers often work in separate silos, speaking different languages. Bridging this gap and unifying their approaches could lead to more efficient collaboration and smoother workflows. Such unification could be facilitated through the development of a code generation platform.


The Benefits of Automation

One of the primary motivations behind building a Codegen platform is the potential to save engineering hours. By automating the translation of designs into code, engineers can focus their efforts on more meaningful tasks, such as building scalable components and addressing complex challenges. Moreover, automation reduces the likelihood of errors, resulting in a more robust and reliable development process.


Building Empathy Through Collaboration

In addition to the practical benefits, a code generation platform can foster greater empathy between designers and engineers. By understanding each other's workflows and challenges, teams can collaborate more effectively and produce better outcomes.


Existing Solutions and Limitations

While there are existing tools and plugins for automating certain aspects of UI development, the talk emphasizes the need for a custom solution tailored to the organization's specific requirements. Existing solutions may not fully align with the organization's framework and may require additional customization.


Introducing Server-driven UI

To lay the groundwork for a code generation platform, the speaker introduces the concept of server-driven UI. Unlike conventional UI development, where components are hard-coded, server-driven UI enables dynamic layout changes based on server responses. This paradigm shift offers numerous advantages, including faster iteration cycles and support for personalization and A/B testing.


Translating Design into Code

The talk demonstrates how design elements in tools like Figma can be translated into code using APIs and intermediate node structures. By extracting meaningful information from design files, engineers can generate models that represent the components in their codebase. These models serve as the foundation for automated code generation.


Generating Code with Custom Templates

Using custom templates and generator functions, engineers can automate the process of translating design models into actual code. By defining special handling for different types of components, such as containers, images, and text, engineers can ensure that the generated code aligns with the organization's framework and standards.


Challenges and Considerations

While a code generation platform offers numerous benefits, it also presents certain challenges and limitations. These include the need to handle unauthorized edits to components, address interactive elements, and support internationalization. Despite these challenges, the potential benefits of a code generation platform are substantial, offering increased efficiency and collaboration in UI development.

Be sure to check out the entire talk and presentation here 🔽

Subscribe to Our Newsletter

More from the engineering frontline.

Dive deep into our research and insights on design, development, and the impact of various trends to businesses.
Insight
Why AI Agents Fail in Production: Building Systems That Recover | Pushkar
Sep 10, 2026

Why AI Agents Fail in Production: Building Systems That Recover | Pushkar

Pushkar’s thegeekconf mini talk explores why AI agents that perform well in demos often struggle in production, and how loud failures, clean context, step monitoring, guardrails, and better agent loops can make them more reliable and predictable.

Insight
Building Is Not Shipping: An AI Agent for App Store Submissions | Gracey Dugar
Sep 8, 2026

Building Is Not Shipping: An AI Agent for App Store Submissions | Gracey Dugar

Gracey Dugar's thegeekconf mini 2026 talk explores why app store submission stalls most AI agent workflows, what an agent needs to finish the job, and a live demo from a single prompt to a published app.

Insight
No More Prompt-and-Wait: Building Autonomous Agents That Act | Kamal Shree
Sep 4, 2026

No More Prompt-and-Wait: Building Autonomous Agents That Act | Kamal Shree

Kamal Shree’s Geekconf mini’s talk explores the shift to autonomous agents, from what separates agents from chatbots, the anatomy and grounding behind them, and how to choose between no-code, low-code, and pro-code build paths.

Insight
From UX to AX: Designing Applications for a World of AI Agents
Sep 4, 2026

From UX to AX: Designing Applications for a World of AI Agents

Learn how Ashita Prasad’s thegeekconf mini 2026 session explores agentic experience and three approaches to building agent-ready applications: Web MCP, MCP Apps, and A2UI.

Insight
Agentic AI: From Copilot to Autopilot | Naveen Kumar Bhansali
Sep 4, 2026

Agentic AI: From Copilot to Autopilot | Naveen Kumar Bhansali

Learn how Naveen Kumar Bhansali’s thegeekconf mini 2026 session explores AI-driven software development, context engineering, enterprise adoption, agentic workflows, and the shift toward building products for AI agents.

Insight
Agentic Commerce: What Happens When Your Agent Tries to Spend Money? | Roopasree Ranganna
Sep 4, 2026

Agentic Commerce: What Happens When Your Agent Tries to Spend Money? | Roopasree Ranganna

Learn how Roopasree Ranganna’s thegeekconf mini 2026 session explores agentic commerce, delegated payments, trust, identity, mandates, and the systems needed to enable AI agents to transact.

Events
 From MVP to Scale: Designing Architecture for AI-First Products
May 11, 2026

 From MVP to Scale: Designing Architecture for AI-First Products

A panel of architects and engineering leaders at thegeekconf mini 2026 discuss how to build and scale AI-first products — from MVP decisions to production-level challenges. The conversation covers data quality, model selection, security, token economics, and the mindset teams need to navigate a fast-moving AI landscape.

The Right Conversation Can

Save You Six Months.

Book a call