#013 • 25 Mar 2025

Contents of this Newsletter

🛠️

Cross-Platform QA

⏰ : 4 min

📖

Recent Blogs

⏰ : 8 min

💻

This Week in AI

⏰ : 3 min

📆

Community Highlights

⏰ : 3 min

MAIN STORY

Can QA Keep Up with Cross-Platform Development?

Cross-platform mobile development has been having its moment. One codebase, two platforms meaning faster releases, lower costs, and less duplication. Flutter and React Native have led the charge, making it easier than ever to build for Android and iOS in one go. Now, the momentum is picking up speed.

ByteDance is shaking things up with Lynx, a Rust-powered framework aiming to put down React Native. Callstack is rolling out an enterprise-grade React Native solution. Recently, a tweet by Fernando Rojo about joining Vercel as Head of Mobile had been making rounds. And now the web giant is rumored to be making a play for mobile. The landscape is shifting. But with speed comes complexity, and QA teams are feeling the pressure.

Fernando Rojo

@FernandoTheRojo

I’ve joined @vercel as Head of Mobile to continue building open source projects for web and native apps. Years in the making. Can’t wait for what’s ahead. Time to ship ▲

8:37 PM • Mar 4, 2025

Retweets

1738

Likes

Read 160 replies

What Is The Catch?

Move fast, break things; but how do you test them? The appeal of cross-platform tools is clear, but the challenges they bring are real:

Same Code, Different Behavior: UI inconsistencies creep in where you least expect them. A button looks perfect in testing but shifts position in production. An animation glides smoothly on an iPhone but stutters on a Samsung due to differences in how each OS renders frames. Snapshot testing catches some of these, but runtime issues often slip through.
Platform-Specific Quirks: Android and iOS don’t just look different, they think differently. A long press behaves one way on iOS, another on Android. Push notifications arrive instantly on one device but get delayed on another due to platform-level handling. Many teams resort to platform-specific workarounds, complicating maintenance and testing.
Performance Trade-Offs: The promise of "write once, run anywhere" falls apart when device fragmentation enters the mix. A cross-platform app that runs flawlessly on a Pixel might drain the battery on a mid-range Android phone. Standard performance testing tools highlight lag but rarely pinpoint whether it’s caused by the framework, native modules, or the platform itself.

Cross-platform development saves time upfront but shifts the burden to QA. Teams often end up maintaining separate test cases for each platform, doubling their workload and negating the supposed efficiency of a shared codebase.

Here’s How We Made It Effortless

At Quash, we knew cross-platform compatibility had to be seamless. We designed our pipeline to make it plug-and-play, ensuring minimal friction for mobile teams.

At Quash, we built our pipeline with modern mobile development in mind. Early on, we encountered legacy methods like Kotlin with XML layouts. But as the industry evolved, declarative UI frameworks like Jetpack Compose became the gold standard for Android development. And then there’s Flutter, where UI and logic live in .dart files. Each evolution in the mobile toolkit led us to refine our approach, ensuring our pipeline stays plug-and-play for today’s development demands.

Here’s how we built our solution:

We first tackled native Android, then adapted for Jetpack Compose, ensuring that both imperative and declarative UI structures were supported.
Flutter was a natural extension, since its UI and logic were structured similarly to declarative UI in native.
The only real challenge was handling launcher files—one for Android, one for iOS. Everything else was plug-and-play.

Our system is built on three key layers:

Unified Data Ingestion: This layer processes various mobile project files, whether they're from legacy Kotlin+XML, modern Jetpack Compose, or Flutter's .dart files. It normalizes these inputs so that our system can work with a consistent dataset.
Knowledge & Orchestration Engine: In this layer, our knowledge graph and workflow combiner analyze the normalized data to map out app structures and user flows. In essence, this layer creates a blueprint of how the app is built, highlighting key components and interactions for testing.
Automated Deployment Pipelines: The final layer routes the processed information into the right testing environments, ensuring that each platform receives tailored and efficient testing.

By adding a dedicated file processing layer at the beginning, we can support different mobile frameworks without overhauling our core testing process. This modular design means our pipeline adapts quickly to new technologies, streamlining QA while keeping our system robust.

When Development Evolves, QA should Too

Cross-platform development and the tools supporting it are evolving fast. But speed without stability is a risk no one should take. At Quash, we’ve built testing to be smart and with AI powered automation we can streamline testing. But edge cases still require human judgement. Automation can handle the majority of the testing, but a human-in-the-loop keeps the quality in check—while validating anomalies, refining edge case handling and more. By doing this, we give you automation that truly understands tests, rather than just running it for you.

Cross-platform apps don’t have to mean double the work for QA. Think about it!

Check out our Recent Blogs

Implementing Microservices: Architecture Patterns and Best Practices

Microservices architecture empowers teams to build scalable, resilient...

Read article

Shift Left Testing: Why Testing Early Is Now Essential

Shift Left Testing is a modern approach that integrates early testing into the Software Development...

Read article

news

This Week in AI

Mistral Small 3.1: A Lightweight Powerhouse

Mistral AI drops Mistral Small 3.1, a 24B open-source model that outpaces Gemma 3 & GPT-4o Mini. With 128k token context, 150 tokens/sec inference, and multimodal capabilities, it’s optimized for on-device use, even on Macs with 32GB RAM.

Google’s Gemini Gets Smarter

Google launches Gemini Canvas & Audio Overview:

Canvas: A real-time AI-assisted coding workspace with live previews.
Audio Overview: AI turns documents into podcast-style summaries with back-and-forth discussions.

OpenAI’s New Responses API & Agents SDK

OpenAI introduces built-in web search, file retrieval, and software automation tools—no external APIs needed. The Agents SDK is now open-source, streamlining AI-driven workflows.

Events

Arali Day 2025: A Gathering of Visionaries

Last Friday, we had the privilege of attending Arali Day at ITC Gardenia, Ashok Nagar, an event that happens once every two years, bringing together a vibrant mix of investors, VCs, and startup founders. As one of our lead investors, Arali Ventures hosted an incredible evening filled with insightful discussions, meaningful connections, and an engaging audience.

Our CEO, Ayush Shrivastava, joined the AI in the Enterprise panel, where the message was clear: enterprises don’t want AI, they want real solutions. The conversation has shifted from “How do we adopt AI?” to “How do we fix QA bottlenecks?”

At Quash, we focus on impact over hype. And we do that by cutting QA time by 80%, not claiming to "revolutionize" development. And when it comes to build vs. buy, even the most technical enterprises prefer buying AI-powered solutions because results matter more than ownership.

Quash was also present with a dedicated stall, where we introduced our product, answered questions, and connected with fellow innovators. The evening wasn’t just about networking, it was a moment to be part of something bigger, a space where bold ideas met real-world execution. And, of course, the food and drinks were exceptional which is a fitting complement to a truly memorable event.

For the Quash team, it was an evening well spent. One that left us inspired and excited for what’s ahead.

Kicking Off The GenAI Collective in Bengaluru

While in San Francisco, our CEO, Ayush Shrivastava, attended events by The GenAI Collective. The experience of networking with AI founders and teams that are building the next wave of AI products truly inspired him and he knew he wanted to bring that energy to Bengaluru.

Last Wednesday, it happened. The first-ever GenAI Collective Meetup brought together AI founders, developers, and enthusiasts for an evening of discussions, demos, and networking. The range of AI-driven products on display was impressive; spanning marketing, finance, travel, legal, and both traditional and no-code development.

The turnout was great, and the conversations were even better. From hands-on AI demos to thought-provoking discussions, the energy in the room was exactly what we hoped for. A huge thanks to everyone who showed up, shared insights, and made this event a success.

And this is just the start. More meetups, more conversations, and more AI-driven innovation are on the way. If you missed this one, stay tuned because we’ll be back soon.

Security | Contact Us | Check out our Blogs | Join Quash Community

Contents of this Newsletter

🛠️

Cross-Platform QA

📖

Recent Blogs

💻

This Week in AI

📆

Community Highlights

MAIN STORY

Can QA Keep Up with Cross-Platform Development?

What Is The Catch?

Here’s How We Made It Effortless

When Development Evolves, QA should Too

Implementing Microservices: Architecture Patterns and Best Practices

Shift Left Testing: Why Testing Early Is Now Essential

news

This Week in AI

​Mistral Small 3.1: A Lightweight Powerhouse

​Google’s Gemini Gets Smarter​

​OpenAI’s New Responses API & Agents SDK​

Events

Arali Day 2025: A Gathering of Visionaries

Kicking Off The GenAI Collective in Bengaluru

Mistral Small 3.1: A Lightweight Powerhouse

Google’s Gemini Gets Smarter

OpenAI’s New Responses API & Agents SDK

Cross-Platform QA Chaos + Google's Gemini & OpenAI's Latest 🧠