Google LLC today announced a host of new, developer-focused updates at its I/O Connect Bengaluru event in India today, including a new integration between Project IDX and Android Studio, so developers can build their apps using a web browser on any device.

It also introduced Firebase AI Monitoring, a new dashboard that shows real-time insights into generative artificial intelligence performance and behavior, and Project Oscar, an AI-agent for maintaining open-source software projects.

In a blog post, Google India’s developer relations lead Karthik Padmanabhan pointed out that India’s community of 4 million developers is at the forefront of AI development today, applying the technology to a wide range of problems, particularly those related to language. Google intends to provide them with all of the resources they need to aid in those efforts.

For instance, the Google DeepMind India team today revealed a new initiative called Project Vaani, which is a collaboration with the Indian Institute of Science that aims to capture the diversity of the hundreds of spoken languages in that country.

The project led to the creation of IndicGenBench, which is a benchmark developers can use to evaluate the capabilities of large language models trained on indic languages. In addition, it has also open-sourced the Composition of Language Models framework, known as CALM, which is for developers who want to combine specialized LLMs with Google’s Gemma models.

In the latest update to Project IDX, online integrated development environment based on Visual Studio Code, Padmanabhan said the company is launching an early preview of Android Studio on Project IDX that can run in any browser. In doing this, developers can quickly fire up a development workspace that runs entirely in their web browser, enabling them to get started with coding, building and testing native Android applications on any device and in any location.

As for Firebase, Google’s mobile and web application development platform, it’s being improved with the introduction of support for the Go programming language in alpha test mode. The idea is to simplify the development of more efficient and scalable Go-based apps by bringing them into the more intuitive Firebase platform, Google explained.

With the preview of the Firebase AI Monitoring dashboard, developers will be able to monitor the performance of their LLMs and the features they support, in production. The dashboard will provide alerts on how those features are being used, their behavior, as well as any inaccuracies and inconsistencies, Padmanabhan said.

Meanwhile, Project Oscar is designed to help open-source project maintainers to keep tabs on the latest status of the projects they run. Padmanabhan pointed out that maintaining open-source projects can be a lot of work, with some of the most popular attracting thousands of contributors who publish hundreds of new commits on a daily basis.

Project Oscar is a reference for an AI Agent that can keep track of all these updates, and is initially targeted at Go, which currently has more than 2,000 contributors and over 93,000 commits. However, Project Oscar can easily be adapted to support other open-source projects, Padmanabhan said.

The Flutter software development kit for creating application user interfaces is also getting a dose of generative AI. The new AI Generated UI feature is able to dynamically compose and personalize UI elements based on natural language inputs, and is launching in early access preview today, the company said.

Finally, Padmanabhan announced that it’s partnering with the Indian Ministry of Electronics and Information Technology to support 10,000 Indian generative AI startups. The support extends to Google Cloud credits, and an updated startup program curriculum to equip founders with the skills and knowledge they need to develop with AI. In addition, Google will also host an annual nationwide hackathon contest to identify India’s next generation of AI innovators.

Image: Google