Google Duet AI makes it easy to gain insights into clustering models and visualize

【Cloud ace cloud oneIt is a global strategic partner of Google Cloud, with more than 20 offices in Asia Pacific, Europe, the Americas, and Africa. Cloud Ace is currently ranked No. 1 in the world for Google Expertise and Expertise, and has won several consecutive Google Cloud awards. As a Google Cloud hosting provider, we offer:Google CloudGoogle MapsGoogle Office SuiteGoogle Cloud Certification Trainingand other services. 】

Get lost in a sea of data

We've all been there: drowning in a sea of data, struggling to navigate complex pipes, and feeling dizzy with data. The frustration of managing a large number of tabs full of different tools and Google searches, and spending a lot of time sifting through data and creating models that meet your needs, can really take the joy out of data discovery.

But is there anything that saves time – maybe even a little fun?It's Duet AI!It treats it as a scientific guide to your personal data, guiding you through any unfamiliar depths of Google Cloud.

New frontiers

Picture this: you see a new dataset full of potential and are given a mission. If you accept, you'll provide your marketing team with 5 sets of leads along with descriptive statistics about them.

Instead of spending a lot of time arguing with tools and doing research, you can use Duet AI to help you.

First, open BigQuery Studio, create a new Python notebook (inside BigQuery Studio) and paste to define the variables, connect to Vertex AI and create a base table.

Next, you start using Duet AI to speed up your development cycle. You prompt Duet for help converting the base table to a BigQuery DataFrame and then displaying the top 10 records.

Draw a circuit diagram

Duet also generates ** for you. All you have to do is check the output and edit it if necessary. Within minutes, the data frame output will appear in your notebook.

Original ** link.

As you dig deeper, Duet AI's generation capabilities will become a valuable asset for you. It seamlessly creates a k-means clustering model that divides your customers into 5 different groups. Duet then writes ** to visually plot these clusters in the form of scatter plots to further facilitate a deeper understanding of the data.

You can add additional ** to aggregate the cluster for descriptive statistics. For example, you can aggregate e-commerce data, such as the average purchase value and number of orders per cluster. These insights met the marketing team's initial requirements.

But you're not done yet!You want to go above and beyond and advise on the next steps your marketing team can take for each customer segment.

Luckily, you can easily reference Vertex AI large language models, such as Text-Bison, directly from BigQuery Studio. You'll soon use the LLM to map out the next steps for your marketing team.

You convert the summary statistics into strings in preparation for LLM calls. You can also define a prompt so that the LLM has the context to generate a customized response.

prompt = f"""you're a creative brand strategist, given the following clusters, come up with \creative brand persona, a catchy title, and next marketing action, \explained step by step.clusters:for each cluster:* title:* persona:* next marketing step:"""

Finally, you prompt Duet AI in your BigQuery Studio notebook to generate a campaign using the prompt variables that you define.

It returns a few things that you think are good, so you can run it.

model = textgenerationmodel.from_pretrained("text-bison")response = model.predict(prompt, max_output_tokens=1024, temperature=0.4)print(response)

It's good that with the help of Duet AI, you've reduced your day's work to minutes.

Set sail

Ready to experience Duet AI for yourselfFollow the ** above to dive deeper into this comprehensive learning path and get hands-on with Duet AI!Contact Cloud Ace and embark on a journey of discovery and unlock the future of data exploration with Duet AI as your data science partner!