How AI advances your DAM: How TESSA works with artificial intelligence

Artificial intelligence (AI) has been a hot topic for years. If you read marketing newsletters, you'll see the term pop up all the time. It often seems as if AI is the universal answer to all marketing challenges. We integrated AI into our TESSA DAM early on. Before we dive deeper, it's worth taking a look at where AI finds relevant applications in digital asset management.

At EIKONA Media, we divide the functions of a DAM into four areas:

• Rule-based asset integration
• Data enrichment
• Easier searchability
• Intelligent distribution of data

As you can see, asset generation has not been our focus so far. Our goal is to provide you with a tool that makes processes more efficient. That is where you will see the greatest benefit.

And AI plays a major role in this context. AI always starts with recognition. Only when a system can reliably recognize objects such as knives, lemons, or colors can you build on it. Without recognition, there can be no generation.

Our claim is clear: a DAM is a process tool. It should relieve you of repetitive tasks. That's why TESSA has been reading metadata from your files since the beginning to make them searchable. If there is an item number in the file name or if it can be assigned elsewhere, you no longer have to manually enter product names into an image. All information from your product information management system (PIM) can be incorporated into the search. This was an early form of algorithmic automation.

The next logical step was AI-supported capture of information that would otherwise have to be laboriously maintained by hand. This was particularly true for mood, ambience, and action images. In the past, these files were often only tagged with locations, photographers, and a few keywords. This was far too little, considering how much detail an image actually contains.

With the Google Cloud Vision API, we were suddenly able to add a large number of additional keywords automatically. Automatically, in multiple languages, and without any extra effort on your part. This opened up new possibilities for marketing teams who needed to find specific images. Today, we are going even further. Instead of assigning keywords, you can search using natural language. You can ask for objects, moods, and compositions. TESSA recognizes images that match your description, even if no single tag has ever been maintained for them. This makes campaign planning much faster. The results are more precise and always up to date, without a team having to re-tag images.

Another advantage is its deep integration with Photoshop and InDesign. This is not just a simple plugin. You can open and edit assets directly from TESSA and incorporate them into documents. We explain how this works in detail in a separate article.

A typical scenario in everyday marketing: You've found an asset that fits perfectly with your new campaign. Maybe it's a product image, a specific perspective, or a particular mood. Now you're looking for variations that fit in seamlessly.

Instead of working with keywords or file names, you use an existing asset as a starting point. The AI analyzes visual features such as shapes, colors, composition, and style and automatically finds other images that match these characteristics. You no longer have to click through folders or laboriously compare motifs.

The advantage is immediately apparent. You find assets that really fit together, whether for product data sheets, multilingual campaigns, or new shop layouts. Visual consistency is maintained and you work much faster. The AI searches your entire DAM in no time and delivers results that would often have remained hidden with a pure keyword search.

As with all features in TESSA, our goal remains the same: AI should help you save time and simplify processes. The similarity search fits seamlessly into your existing workflow. You select an image, start the search, and immediately receive a carefully curated selection of matching assets. There's hardly a more efficient way to make visual decisions.

Asset management is only a small part of a DAM, but it is an important one. After all, a DAM should free you from repetitive, monotonous work. Anything that sounds like batch processing can be handled by TESSA. It may not be highly complex AI, but it is extremely practical.

For years, TESSA has been supporting automated processes and channels, which we have already described elsewhere. What has been added in the meantime are AI tools that we connect via API. This allows you, for example, to automatically remove backgrounds in product images. You select the desired assets, click on Remove Background, and the image is cropped. You can still work out the final details in Photoshop if necessary. In any case, you'll be significantly faster.

Here, too, our principle applies: we don't use just any tool, but the best on the market. We focus on our core business, and you work with the most powerful AI tools available. It feels a bit like microservices. TESSA is at the center of your asset work and specifically fetches the functions you need. This allows you to use the complete feature set of Adobe Photoshop. Without copying back and forth, your assets remain available in TESSA DAM.

One thing is clear: TESSA does not generate assets. The focus is on bringing assets into the DAM cleanly, enriching them with metadata, making them findable, and reliably outputting them to any target system.

However, if you already use tools that generate assets using AI, we can easily integrate these systems. It doesn't matter whether these are the well-known AI tools you regularly see on LinkedIn or professional solutions that generate photorealistic images and videos from 3D models. As soon as assets are created, TESSA ensures that they are integrated into your workflow.