The first half of 2025 has been transformative for Google Cloud, delivering critical updates across AI, data infrastructure, application development, and analytics. These innovations are not just global milestones, they hold immediate potential for businesses here in Thailand, from startups to large enterprises.

At Digigen, we keep a close eye on what’s changing and what’s coming. Here's our curated look at the most important announcements and what they mean for Thai organizations navigating digital transformation.

🌟 Gemini 2.5: The Smartest Yet

Gemini 2.5, Google’s most capable foundation model to date, is now generally available via Vertex AI. With it comes significant enhancements in natural language processing, image and video understanding, and advanced code generation.

In practical terms, Gemini 2.5 enables:

• Rapid development of customer-facing chatbots and agents

• More nuanced content summarization and generation

• Enhanced document processing across multiple languages

For businesses in Thailand, especially those operating in bilingual environments or handling large volumes of customer interaction, Gemini unlocks a new level of productivity and automation. Early adopters are integrating it into customer service flows, compliance checks, and even internal training content.

📊 Serverless Spark Goes GA in BigQuery

A major step forward for data engineering: Serverless Spark is now fully integrated with BigQuery. This unified experience means teams can run SQL and Spark workloads side-by-side on the same data, no complex pipelines, no switching environments.

This is particularly valuable for Thai organizations looking to reduce data latency while supporting diverse data teams (SQL analysts + Spark developers). With up to 3.6x performance improvements thanks to the Lightning Engine, we expect broader adoption in sectors like fintech, retail, and telecom.

⚡ Pub/Sub SMTs: Streaming Intelligence at the Edge

Google Cloud Pub/Sub has introduced Single Message Transforms (SMTs) — lightweight, real-time data transformations using JavaScript UDFs. You can now enrich or clean individual messages as they move, reducing reliance on intermediate processing tools.

What this means for Thailand: organizations dealing with logistics, real-time monitoring, or IoT can simplify their architecture and enhance responsiveness — especially important in Bangkok’s smart city initiatives or in energy infrastructure monitoring.

🎥 Cloud Run with GPU Support

Cloud Run now supports NVIDIA GPUs, allowing for efficient, serverless deployment of compute-heavy workloads. From AI video generation to real-time 3D rendering, this unlocks use cases previously reserved for highly specialized infrastructure.

We see this as a step-change for creative studios, architecture firms, and research institutions in Thailand that want access to high-performance compute without the headache of managing servers.

🧬 Datastream for MongoDB

With MongoDB now supported as a source in Datastream, it's easier than ever to move data from transactional systems to analytics platforms like BigQuery in real time. This supports modern use cases where instant visibility into app data is essential — from e-commerce dashboards to logistics control towers.

The ability to do this with minimal engineering work means smaller teams can now adopt enterprise-grade data architectures.

🎬 Scene Detection in Video with Gemini

Gemini's scene transition detection is a small but fascinating breakthrough in video AI. It moves beyond just recognizing hard cuts to understanding narrative structure, using multimodal input (video, audio, and text).

Media companies, content marketers, and digital agencies in Thailand can start to use this for:

• Automatically generating highlight reels

• Indexing long-form content

• Aligning video ads with narrative context

It’s a great example of AI turning content into structured, actionable data.

🤖 AI Agents: Now Easier to Deploy

Google Cloud introduced tighter integration between AI Studio and Cloud Run, plus support for the Model Context Protocol (MCP). This means developers can deploy multi-agent systems with far fewer steps.

Agent-based automation is starting to emerge in HR, finance, and customer experience teams across Thailand. With these tools, you no longer need a huge DevOps team to run secure, scalable AI agents.

🧩 Broader Ecosystem Highlights

Beyond product launches, Google Cloud has made several ecosystem changes worth noting:

• Marketplace reforms to give partners better margins and flexibility

• Expanded support for Oracle workloads and cross-cloud disaster recovery

• Enhanced database monitoring with Gemini-powered insights in the new Database Center

These improvements are especially beneficial for organizations with hybrid architectures — a common setup among larger Thai enterprises.

🌏 What It All Means for Thailand

Thailand’s digital economy is growing rapidly, and Google Cloud’s latest updates are helping local businesses scale more intelligently. Whether it’s integrating AI, simplifying infrastructure, or building real-time apps, the message is clear: these tools are more accessible than ever.

At Digigen, our focus is on guiding Thai businesses through this journey — with localized contracts, billing in THB, and deep technical enablement.

We’re excited about what H2 2025 will bring. For now, there’s already more than enough to build what’s next.

📨 Want to explore these updates hands-on?

Digigen is here to help you get the most out of Google Cloud in Thailand.

➜ Talk with our consultants at Digigen, your trusted Google cloud partner in Thailand.