Hello from Surya Digital!

Over the past few months, we’ve been working with a large e-commerce platform serving hundreds of thousands of active users to enhance its Core Web Vitals — Largest Contentful Paint ( LCP), Cumulative Layout Shift (CLS), and Total Blocking Time (TBT) — through a complete restructuring of its frontend architecture for performance.

Key improvements included:

• Restructuring the architecture to better leverage server-side rendering (SSR) and code-splitting, ensuring faster initial loads and reduced TBT.
• Using React Suspense and dynamic imports so that independent sections load in parallel, improving both LCP and TBT.
• Lazy-loading analytics scripts and minimising third-party injections to reduce main-thread blocking.
• Replacing heavy third-party UI libraries with lighter, SSR-compatible alternatives.
• Removing full icon libraries and shipping only the required icons to cut down unused CSS and JS.
• Statically generating low-change pages for near-instant loads.
• Adding shimmer placeholders to reduce layout shifts and improve CLS.

The result — a leaner, more scalable architecture that delivers faster load times and a smoother user experience.

Under the Hood

AI for Junior Software Engineers

Are you holding your junior developers back by letting them use AI? Read our founder Gopal’s strong take on why we ban AI tools for engineers with less than a year of experience.

Sketching & Writing Things Down Reduce My Anxiety as a Product Designer

Want to reduce product design anxiety? Our resident UI expert, Sai explains why paper and pen are still the most powerful tools for clarifying your ideas and solving the right problems.

Breakroom Buzz

Debugging Minimal Containers: Why We Switched to docker debug

We've made the switch: docker debug (introduced in 2024) is now our primary tool for troubleshooting minimal containers, replacing the old standby docker exec -it.

The old method fails when slim images lack shells or utility packages. docker debug is a game-changer because it seamlessly injects a full, customizable toolbox (like strace, tcpdump, and shells) into the running container for deep diagnostics. This avoids ever having to bloat your images.

We've incorporated this approach into our training. Is it part of your routine yet?

Cursor's New Coding Model and Agent Interface 

The Cursor AI-first code editor just launched its massive 2.0 update! The headliners are the new, 4x faster Composer model and Multi-Agents, which let you run up to eight AI agents in parallel. The update also brings Voice Mode for hands-free control and makes its in-editor browser and secure sandboxed terminals generally available. If you use Cursor, this is a must-see!

Off The Assembly Line

Last week, we hosted an 8-hour workshop with Subject Matter Experts from Avenza Consulting Services, a Surya group company specializing in providing consulting services around Temenos Transact core banking system to some of the biggest global banks. Using 5 complex BRDs, Lidana generated 70% of the required test cases in one go and achieved 100% coverage in just 13 minutes, cutting manual effort by over 90%.

Curious how we did it? Read more here.

If you'd like to experience how Lidana can streamline test definition and accelerate your quality assurance efforts, find out more at lidana.ai, or reply to this email.

We’ll see you in the next edition with more stories and updates. Until then, happy innovating!