Datacadabra's roadmap
In our previous blog, we talked about perception as the first of the 4 steps in Datacadabra's AI roadmap. In this blog, we explain how to process, learn to understand and structure the information you get from data. We do that structuring using a Digital Intelligence Framework, which we'll abbreviate to DIF for convenience. This DIF is an integrated set of technologies and tools that enables Datacadabra to effectively develop, implement and manage digital intelligence, such as artificial intelligence (AI) and machine learning (ML).
DIF as a block box
Such a DIF is best compared to a block box. In that block box, you have different blocks with different shapes and colors. Each block has a specific functionality in the digital intelligence framework, such as data management, model development, monitoring, et cetera.
Before you start building, you need a build plan. This corresponds to determining the goals and scope of your AI project before you deploy the DIF. Depending on your building plan, you choose specific blocks from the block box. In the digital intelligence platform, these are the tools and services you select based on your project needs, such as data storage or machine learning frameworks.
Combining tools and services into a model
Now you are ready to start stacking the selected blocks to add structures. In the DIF, you build an AI model by combining different tools and services for model development. Sometimes the blocks need to be modified to fit together better. This is equivalent to performing data preprocessing and feature engineering to prepare your data for the model. If you decide to expand your build, you add more blocks. In the digital intelligence platform, you can scale by adding additional services.
Animal with stripes
Are you still here? We can imagine that this is a somewhat technical story. Therefore, let's illustrate it with an example. In our previous blog, we cited the example of a picture of an animal with stripes. From that image, one might conclude that it is a zebra. But to make a definitive statement about this, you need to create context so that you can further process and learn to understand the information you see. Other information may also need to be added. After all, it could also be a tiger, also an animal with stripes. By further filtering the information, you can make statements about what you see in the image with increasingly accurate precision. From the filtered information, you can now also make a statement about the color stripes (black and white), but discover from other features (a tail and mane) that you are dealing with an animal. So the probability is quite high that it is a zebra.
So processing and structuring information is what we do in Step 2 of our AI roadmap. In our next blog, we'd like to take you to step 3, which focuses on understanding information.
The white paper DIF in your mailbox?
Datacadabra has created a white paper in which we explain how the DIF works. Using an example, we take you step-by-step through the process of structuring and understanding data perception so that models can be trained on it.
Curious about the whitepaper? Fill in your details below and you will receive our white paper on the DIF in your email
