Businesses are investing in tools and talent to facilitate data-driven decision-making. Machine learning models and data science techniques can give organizations an opportunity to make sense of the vast amount of data they’ve collected.
However, many smaller companies are struggling to adapt to this trend because of limited resources and a lack of knowledge exports within their team.
In this article, we will take a look at Obviously AI, an exciting new AI no-code tool that aims to help organizations build predictive models that will drive growth.
We’ll cover some of the key features Obviously AI provides and walk you through how to start using the tool to create models in minutes.
What is Obviously AI?
Obviously AI is a no-code data science platform that allows users to create predictive models using their own data. The platform uses an intuitive user-friendly interface that allows you to upload or connect your data sources and build models in just a few minutes.
Once your data has been uploaded to Obviously AI’s servers, you’ll be able to explore your dataset to get a more comprehensive view of what you know and what you would like to predict.
Obviously AI will then use your dataset to create machine learning models that you can use right away. These models can then be integrated with third-party tools such as Zapier or through their own REST API service.
Obviously AI also offers a user-friendly dashboard interface for your model, allowing anyone to use the model right from their browser. For example, the image above shows a user adjusting the details of a customer to predict their churn rate or the probability that they’ll unsubscribe from a service.
All these tools become even more powerful with the ability to consult data scientists to help users understand how to best use their data.
Key Features of Obviously AI
Here are some of the key features available in Obviously AI.
Obviously AI comes with an easy-to-read data explorer that will help you make sense of your uploaded datasets. Users can see the total size of the dataset as well as the value distribution of each column.
Obviously AI will also analyze your datasets and will let you know which fields are missing data and which fields are ready to use for building a model.
Support for Multiple Models
Obviously AI offers support for models that handle clustering, classification, time-series analysis, and regression.
The platform will test several different models and return the model with the best overall result.
Once Obviously AI build a model, users will be able to send requests to a custom REST API.
This will allow organizations to use Obviously AI as a back-end cloud service that can classify or predict any type of industry data they encounter.
The platform even includes support for Zapier, allowing users to integrate Obviously AI’s models into their workflow automations.
Data Scientist as a Service
Since most organizations start off with raw unfiltered data, it may be necessary for them to consult a professional data scientist for help.
Obviously AI understands this need and added several features to the platform to accommodate this.
Users who are subscribed to the Software + Data Scientist plans will be able to work with data scientists to help make more accurate models.
Data scientists can look into your raw datasets and create new columns, handle missing data, and identify the best models for your use case.
How to Build a Model in Obviously AI
Now that you have a better idea of what the Obiously AI platform has to offer, let’s take a look at how to build an Obviously AI model.
In this quick example, we will use Obviously AI to create a classification model using the Titanic Survivors dataset.
First, log in or create an account with Obviously AI. Access the website app.obviously.ai to view the Obviously AI dashboard.
Click on the “+” button on the left panel to add a new dataset to your account.
Select the type of model you want to build. In this demonstration, we will select Classification.
Next, choose how to upload your dataset. For now, we’ll upload our data by uploading a local CSV file.
The Review page will give you a preliminary overview of your data. You’ll also see what percentage of each column is empty and ready to use.
In the next page, you must decide which column to predict. In this case, we’ll choose the Survived column.
Once the model has been built, you’ll be redirected to a new view containing a summary of the model.
Obviously AI will inform you what type of model was used as well as the accuracy of the model. The page will also list the features in your dataset that are the top drivers for predicting the result.
Click on the Use Model tab and select Launch Model to try out the model yourself in your browser.
Obviously AI will redirect you to a dashboard where you can input data for prediction. In the example above, we’ve entered data of a new passenger. Clicking on Predict, we’ve determined that there is a 78% probability that the new passenger survived (Survived=1).
We can also make batch predictions by uploading a CSV of test data.
Click on Download Predictions to download a CSV of the predictions the model made on the uploaded data.
Obviously AI offers a free plan with a limit of 1200 predictions and access to classification and regression models.
The Limited Access plan increases the amount of predictions to 12,000 and gives users access to a dedicated data scientist to help with their projects.
The Full Access plan increases the limit to 120,000 predictions and unlimited rows of training data. Users will also have access to time series and clustering models and a dedicated Data science team to handle data preparation and other requests.
After using Obviously AI to create my first model, I’m impressed with how straight-forward and quick it is to convert a raw data set into a working cloud-based data science application. It takes just a few minutes to set up a model from scratch.
While the pricing plans may be a bit steep for hobbyists or non-profit organizations, the plans are reasonable enough for smaller companies that do not yet have a dedicated data scientist or data engineer on their team.
What do you think of Obviously AI?