Google operates the free data science agent that works in Jimini on the Clap Bethon platform

Artificial intelligence agents are all anger, but what about one who focuses in particular on analysis, sorting and drawing conclusions from wide volumes of data?

Google Data science agent Is this only: the new Gemini 2.0 assistant that works of artificial intelligence and who works to automate data analysis is now available for users between 18 years old in countries and free languages.

The assistant is available through Google Colab, the eight -year -old company service to run Python Code Live Online at the top Tensioner processing units (TPUS).

It was initially launched for the trusted laboratories in December 2024, the data science agent was designed to help researchers, data scientists and developers to simplify the functioning of their work by generating the entire Jupyter books of natural language descriptions, all in the user browser.

This expansion is in line with Google’s ongoing efforts to integrate coding features and data sciences that AI move in Colab, and build on previous updates such as helping to develop a working artificial intelligence, which has been announced in May 2023.

It also works as a kind of advanced and late Rejonder to Openai’s Advanced data analysis Chatgpt (previously Translator), Which is now combined into ChatGPT when running GPT-4.

What is Google Colab?

Google Colab (Colabook) is the JUPYTER notebook based on the view that enables users to write and implement the Python icon directly in their browser.

Jupyter Notebook is an open source web app that enables users to create and share documents that contain direct software instructions, equations, perceptions and narrative text. She grew up from the iPython project in 2014, and now supports more than 40 programming languages, including Python, R and Julia. This interactive platform is widely used in data science, research and education for tasks such as data analysis and visualization of programming concepts.

Since its launch in 2017, Google Colab has become one of the most widely used platforms for machinery ML).

As Ori Abramovsky, data science at Specialops.IO, detailed in Excellent medium function Since 2023, it has made the ease of use of colum and free access to graphics processing units and TPUS is a prominent choice for many developers and researchers.

He pointed out that the low barrier to entry, smooth complementarity with Google Drive and TPUS support allowed his team to shorten training courses significantly while working on artificial intelligence models.

However, Abramovsky also referred to the restrictions of Colab, such as:

• The boundaries of the session (Especially for free -class users).
• Unpredictable resource customization At peak use times.
• Lack of critical featuresSuch as implementing the effective pipeline and advanced scope.
• Support ChallengesGoogle provides limited options for direct help.

Despite these defects, Abramovsky stressed that Colab is still one of the best notes notebook solutions – especially in the early stages of data analysis projects and data projects.

Simplify data analysis with artificial intelligence

Data science agent is based on the Clamp server notebook by eliminating the need for manual preparation.

Using Google Gemini Ai, users can describe their analytical goals in simple English (“Imagine trends,” “Prediction Model Training,” “Clean valuable values”The agent generates a fully implemented Colab books.

Supports users through:

• Automation of the analysisHe generates full work notes instead of isolated code scraps.
• Time savingsIt eliminates manual preparation and frequent coding.
• Promoting cooperation: Features of the combined participation features of collective projects.
• Provide adjustable solutionsUsers can adjust and customize a created code.

Data science agent actually accelerates scientific research in the real world

According to Google, I reported the first laboratories of a large time provision when using the data science agent.

For example, a scientist at the Lawrence Berkeley National Laboratory, which works on methane emissions in tropical lands, estimated that the time of data processing has decreased from one week to only five minutes when using the agent.

The tool also performed well in industry standards, as it ranked fourth on Dabstep: Data Agent Standard for Multi -Step In embracingBefore artificial intelligence factors such as React (GPT-4.0), Deepseek, Claude 3.5 Haiku and Llama 3.3 70B.

However, OPENAI models compete with O3-MINI and O1, as well as Claude 3.5 Sonnet of humans, both exceeded the new Gemini Data Data agent.

Start

Users can start using the Google Colab Data agent by following these steps:

1. Open a new colum notebook.
2. Download a collection of data (CSV, json, etc.).
3. Naturally analysis row Using the side board.
4. Implementing the notebook created To see visions and perceptions.

Google provides a sample of data collections and fast ideas to help users explore its capabilities, including:

• Overflow developer survey: “Imagine the most popular programming languages.”
• Iderrate types of data collection: “Person, Speerman, and Cindle, the links.”
• Glass classification data collection: “Random Forest Basete.”

At any time the user wants to use the new agent, they will have to move to Colab and click “File”, then “New notes in the drive”, and the resulting notebook will be stored in their Google Drive Cloud account.

My brief experimental use was more mixed

I was granted, I am a modest technical journalist and not a data world, but my own use of the new Data Agency Gemini 2.0 in Colab has so far been less smooth.

You have downloaded five CSV files (separated values ​​and standard script files from Excel or Sheets) “How much I spent every month and a quarter on my companions?”.

Moving forward and performed the following operations:

• Complied data setsTreat the history and disparities of the account number.
• It was nominated and the data is cleanedEnsure relevant expenses only.
• Collected transactions In the month and a quarter to calculate spending.
• Disquies bornLike the line of directional analysis.
• The summary results In a clear and organized report.

Before implementation, Colab paid a confirmation message, reminding me that it may interact with external application programming facades.

He did all this very quickly and smoothly in the browser, within seconds. It was impressive to watch her through analysis and programming with visual descriptions step by step for what she was doing.

However, in the end, it led to the creation of an inaccurate graph that shows interest spending for only one month, failed to recognize the papers, and included a full year’s value that was broken months ago. When I asked for this, I tried wonderfully, but in the end I could not produce the correct software instructions series to answer my demands.

From the zero point, I tried exactly the same router on a new book in Google Colab, and produced a much better result, but it is still strange.

I will have to try to explore and fix errors in each other, and as I said, the initial wrong result may be due to a lack of experience in using data science tools.

Colat pricing and artificial intelligence features

Although Google Colab is still free, users who need additional mathematical energy can upgrade to paid plans:

• Collap Pro ($ 9.99/month): 100 account units, faster graphics processing units, more memory, and peripheral access.
• Collap Pro+ ($ 49.99/month): 500 account units, priority GPU promotions, background implementation.
• Colab EnterpriseGoogle Cloud Integration, Generation Code-Code.
• Pay as you-G9.99 dollars for 100 arithmetic units, $ 49.99 for 500 arithmetic units.

In addition to the data science agent, Google expands the capabilities of artificial intelligence within the colum.

Google collects claims, the symbol created and user comments to improve artificial intelligence models. While the data is stored for up to 18 months, it is unknown, and the deletion requests may not always be fulfilled. Users are advised not to provide sensitive or personal information, as human auditors may process claims. In addition, the code that is created from artificial intelligence should be reviewed carefully, as it may contain inaccuracy.

Reactions, welcome

Google encourages users to make notes through the Discord Google Labs community on the #-Dative-OGENT #.

Although the automation driven by artificial intelligence has become a major trend in data science, the Google Data Agent in Colab can help researchers and developers to focus more on ideas and less on coding. With the tool expanding to more users and regions, it will be interesting to see how the future of analyzes that are with the help of AI.