Google Launches Gemini-Powered Data Science Agent on Colab Platform

Google Releases Data Science Agent in Colab

The agent achieves goals set by the user by orchestrating a composite flow which mimics the workflow of a typical data scientist. Google released a Data Science Agent on the Colab platform on Monday, powered by its Gemini 2.0 AI model. The Data Science Agent is capable of autonomously generating the required analysis of the data file uploaded by the user. It is also capable of creating fully functional notebooks, and not just code snippets. Google said the agent "removes tedious setup tasks like importing libraries, loading data, and writing boilerplate code". The agent achieves goals set by the user by "orchestrating a composite flow" which mimics the workflow of a typical data scientist. Users can use the agent to clean data, perform exploratory data analysis, statistical analysis, predictive modeling and other such tasks. The generated code can be customised and extended to meet users' needs. Moreover, results can also be shared with other developers on Colab. Google also said that the agent ranked fourth on the DAPStep (Data Agent Benchmark) on HuggingFace, ahead of GPT-4o, DeepSeek-V3, Llama 3.3 70B and more. The Data Science Agent was launched for trusted testers last December, but is now available on Google Colab. Colab is a free, cloud-based environment where Python code can be written and run within the web browser. It also provides free access to Google Cloud GPUs and TPUs. "We want to simplify and automate common data science tasks like predictive modelling, data preprocessing, and visualisation," Google said. Recently, Google also announced the public preview of Gemini Code Assist, a free AI-powered coding assistant for individuals. The tool is globally available and supports all programming languages in the public domain. It is available in Visual Studio (VS) Code and JetBrains IDEs, as well as in Firebase and Android Studio. Google also said the AI coding assistant offers "practically unlimited capacity with up to 1,80,000 code completions per month".

Google upgrades Colab with an AI agent tool | TechCrunch

Google Colab, Google's cloud-based notebook tool for coding, data science, and AI, is gaining a new "AI agent" tool, Data Science Agent, to help Colab users quickly clean data, visualize trends, and get insights on their uploaded data sets. First announced at Google's I/O developer conference early last year, Data Science Agent was initially launched as a standalone project. However, Google decided to integrate it into Colab with the goal of helping users access the agent directly from a Colab notebook, said Kathy Korevec, director of product at Google Labs, in an interview. Data Science Agent is available for free as of this week in Colab, although Colab limits free users to a relatively low amount of computing. Google offers a range of paid Colab plans with higher limits starting at $9.99. Data Science Agent is primarily aimed at data scientists and AI use cases, but the agent can also help find API anomalies, analyze customer data, and write SQL code. All users need to do is upload their data and ask the agent a question. Data Science Agent uses Google's Gemini 2.0 AI model family on the backend, along with "reasoning" tools to help with feature engineering and data cleaning tasks. Korevec told TechCrunch that Google is constantly improving the agent and using techniques including reinforcement learning, as well as integrating user suggestions, to enhance Data Science Agent's performance. Data Science Agent currently only supports CSV, JSON, or .txt files under 1GB in size. It can analyze about 120,000 tokens in a single prompt, which works out to about 480,000 words. Korevec said that Data Science Agent may come to additional dev-focused Google apps and services in the future. "We're scratching the surface of what people can do here," she said. "Because it's an agent, we can integrate it into a bunch of different tools, and I don't necessarily want to force people who are shy about looking at the code to go to Colab."

Google launches free Gemini-powered Data Science Agent on its Colab Python platform

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More AI agents are all the rage, but how about one focused specifically on analyzing, sorting, and drawing conclusions from vast volumes of data? Today, Google announced that its Data Science Agent, a new, free AI assistant powered by its model Gemini 2.0 for automating data analysis, is now available to users aged 18+ in select countries and languages for free. It's available through Google Colab, the company's eight-year-old service for running Python code live online atop the graphics processing units (GPUs) owned by the search giant and its own, in-house Tensor Processing Units (TPUs). Initially launched for trusted testers in December 2024, Data Science Agent is designed to help researchers, data scientists, and developers streamline their workflows by generating fully functional Jupyter notebooks from natural language descriptions, all in the user's browser. This expansion aligns with Google's ongoing efforts to integrate AI-driven coding and data science features into Colab, building on past updates such as Codey-powered AI coding assistance, announced in May 2023. It also acts as a kind of advanced and belated rejoinder to OpenAI's ChatGPT Advanced Data Analysis (previously known as Code Interpreter), which is now built into ChatGPT when running GPT-4. What is Google Colab? Google Colab (short for Colaboratory) is a cloud-based Jupyter Notebook environment that enables users to write and execute Python code directly in their browser. Jupyter Notebook is an open-source web application that enables users to create and share documents containing live code, equations, visualizations, and narrative text. Originating from the IPython project in 2014, it now supports over 40 programming languages, including Python, R, and Julia. This interactive platform is widely used in data science, research, and education for tasks like data analysis, visualization, and teaching programming concepts. Since its launch in 2017, Google's version, Google Colab, has become one of the most widely used platforms for machine learning, data science, and education. As Ori Abramovsky, Data Science Lead at Spectralops.io, detailed in an excellent Medium post from 2023, Colab's ease of use and free access to GPUs and TPUs make it a standout option for many developers and researchers. He noted that the low barrier to entry, seamless integration with Google Drive, and support for TPUs allowed his team to dramatically shorten training cycles while working on AI models. However, Abramovsky also pointed out Colab's limitations, such as: * Lack of critical features, like efficient pipeline execution and advanced scheduling. * Support challenges, as Google provides limited options for direct assistance. Despite these drawbacks, Abramovsky emphasized that Colab remains one of the best serverless notebook solutions available -- particularly in the early stages of machine learning and data analysis projects. Simplifying data analysis with AI The Data Science Agent builds on Colab's serverless notebook environment by eliminating the need for manual setup. Using Google's Gemini AI, users can describe their analytical goals in plain English (e.g., "Visualize trends," "Train a prediction model," or "Clean missing values"), and the agent generates fully executable Colab notebooks in response. It offers the following features: * Automates analysis - Generates complete, working notebooks instead of isolated code snippets. * Saves time - Eliminates manual setup and repetitive coding. * Enhances collaboration - Uses Colab's built-in sharing features for team-based projects. * Offers modifiable solutions - Users can adjust and customize the generated code. Already accelerating real-world scientific research According to Google, early testers have reported significant time savings when using Data Science Agent. A scientist at Lawrence Berkeley National Laboratory, working on tropical wetland methane emissions, estimated that their data processing time dropped from one week to just five minutes using the agent. The tool has also performed well in industry benchmarks, ranking 4th on the DABStep: Data Agent Benchmark for Multi-step Reasoning on Hugging Face, ahead of AI agents such as ReAct (GPT-4.0), Deepseek, Claude 3.5 Haiku, and Llama 3.3 70B. However, OpenAI's rival o3-mini and o1 models, as well as Anthropic's Claude 3.5 Sonnet, both outclassed the new Gemini Data Science Agent. Getting started Users can start using Data Science Agent in Google Colab by following these steps: Google provides sample datasets and prompt ideas to help users explore its capabilities, including: * Stack Overflow Developer Survey - "Visualize most popular programming languages" * Iris Species dataset - "Calculate and visualize Pearson, Spearman, and Kendall correlations" Granted I'm a lowly tech journalist and not a data scientist, but my own usage of the new Gemini 2.0-powered Data Science Agent in Colab so far has been less than seamless. I uploaded five CSV files (comma separated values, standard spreadsheet files from Excel or Sheets) and asked it "How much am I spending each month and quarter on my utilities?". The agent went ahead and performed the following operations: * Merged datasets, handling date and account number inconsistencies. * Filtered and cleaned the data, ensuring only relevant expenses remained. Before execution, Colab prompted a confirmation message, reminding me that it might interact with external APIs. It did all this very rapidly and smoothly in the browser, in a matter of seconds. And it was impressive to watch it work through the analysis and programming with visible step-by-step descriptions of what it was doing. However, it ultimately generated an inaccurate graph showing just one month's utility spending, failing to recognize the sheets included a full year's worth broken out by months. When I asked it to revise, it gamely tried, but ultimately couldn't produce the correct code string to answer my prompt. I tried from scratch with the exact same prompt on a new Notebook in Google Colab, and it produced a far better, yet still odd result. I'll have to try troubleshooting it some more, and as I said, the initial erroneous result may be due to my own lack of experience using data science tools. Colab pricing and AI features While Google Colab remains free, users who need additional compute power can upgrade to paid plans: * Colab Pro ($9.99/month) - 100 compute units, faster GPUs, more memory, terminal access. * Colab Pro+ ($49.99/month) - 500 compute units, priority GPU upgrades, background execution. In addition to Data Science Agent, Google has been expanding AI capabilities within Colab. Google collects prompts, generated code, and user feedback to improve its AI models. While data is stored for up to 18 months, it is anonymized, and deletion requests may not always be fulfilled. Users are advised not to submit sensitive or personal information, as human reviewers may process prompts. Additionally, AI-generated code should be reviewed carefully, as it may contain inaccuracies. Feedback welcome Google encourages users to provide feedback on Data Science Agent through the Google Labs Discord community in the #data-science-agent channel. With AI-driven automation becoming a key trend in data science, Google's Data Science Agent in Colab could help researchers and developers focus more on insights and less on coding setup. As the tool expands to more users and regions, it will be interesting to see how it shapes the future of AI-assisted analytics.