We are experiencing an explosion of AI technologies, rapidly moving from research to becoming commodity technologies. During this past year, one of the technologies that has attracted the most attention is in the field of generative AI technologies [5], [6]. More precisely, the technology that can generate synthetic images, art or simply modifying the style of pictures and that is available in applications such as DALL-E [7], Nightcafe [8], Midjourney [9], and others [10]. A simple text description of our image including the style can instantly create synthetic images very close to real counterparts or even pieces of art (text-to-image [11], [12]). Not only that, but it can also transform the style of our real images and retouch certain parts of the image. This is just a case within generative AI technologies. AI applications such as automatic generation of texts, music, art, faces, animations, restoration of old movies and aging or rejuvenation of faces are other examples within this category. Other types of AI technologies include automatic communication through increasingly sophisticated chatbots (Conversational AI [13]), automatic translation systems in a multitude of language pairs, both written and spoken (Machine Translation [14]), recommendation systems [15], natural language systems to extract key sentences or entities (NLP AI [16]), automatic recognition of objects, content and people in images or video (Computer Vision AI [17]) and many others. These AI technologies are expected to fully handle tasks related to content creation and generation of images, videos, and audio. Deal with the automation of processes, the search, recovery and synthesis of stored information, the recognition of people and writing, with human-machine communication but also with more intellectual complex activities within the field of content and report generation and the extraction of insights of high value for companies.
But this is just the beginning. We will be completely astonished at what these technologies will be able to do for us in the coming years. The most important thing is that these technologies go from being experimental to being part of everyday use in our lives and businesses in a very short time. These technologies are not here to replace us but to help us in boring and repetitive tasks, in tasks where the human performs cognitive activities that do not involve any effort for us but are present in our daily lives and businesses. All of them, will generate a huge number of opportunities for the full digitization of businesses and consequently will result in great savings and the optimization of business processes [18], [19]. Recent studies show for instance, that AI-powered call agents will replace humans in 15% of customer communications in the next three years, which implies an estimated saving of $80B [20]. Many of the AI technologies mentioned above are already being exploited as Cognitive Services in the cloud, through REST API or by other means, quickly becoming commodities and being able to be used by ordinary people and companies.
But how can I incorporate these AI technologies in my company? How can all these cognitive services and AI help me in my business, in new customer needs and in process optimization? The answer is that many companies already have the data and access to the necessary technologies to use these cognitive AI services in supporting their businesses. Many companies began a long time ago to accumulate data and adopt Big Data technologies. Much of this data remains stored in Data Lakes or similar digital storages without being explored or exploited. There is great potential in these sources using Big Data Analytics and AI Cognitive Services technologies. A single source or a combination of several ones is enough to carry out many use cases and optimize processes. For instance, same streaming video from on-site cameras could serve to different purposes. It can be reused for automatic intruder detection, safety of people in dangerous environments, automatic reporting of industrial activities and inventories, production triage, and defect detection. In these cases, the computer vision cognitive services would perform most of the tasks. As a further example, a single source of documents, texts or written communications can be utilized for automatic document classification, document association, summarization and extraction of important ideas or keywords. It can also be used as a source of information in smart searches. In all these cases, AI cognitive services for language will be of great help in developing solutions.
And how well are companies doing today in relation to AI technologies and Cognitive Services? Are firms using all these technologies to really improve their business and their relationships with customers? The reality is that many companies lack a consistent vision and a strategy regarding the use of Big Data Analytics, AI, and cognitive services. Many firms have simply failed or are now struggling to maintain the use of such technologies in their business [21], [22], [23]. The reasons why many companies have not been successful in the use of AI and Big Data technologies are diverse. One of the explanations is that many companies focus on resolving isolated cases or at most several of a similar nature. Companies develop data pipelines and AI models from scratch and in most cases using open-source technologies. Hence, the effort to adapt their framework to more recent AI-technologies or new data, updating AI models and maintaining the infrastructure is that big that it makes it unsustainable in the long term. Other causes have to do with that on many occasions neither the data nor the data preprocessing were adequate to the business case. Recent studies also show that more than half of the companies (58%) focus on resolving the so-called "need-to-do" cases and only 46% of the firms carry out the "must-dos" that provide great benefits with little complexity [24].
There is also the problem of trying to build use cases around data that is of analog origin instead of redesigning systems to use purely digital data. Say for instance, handwriting texts, scanned documents, or forms filled out by hand. It is also important to consider that all these AI technologies evolve very quickly, which demands highly specialized professionals, up-to-date and technical knowledge, as well as the continuous updating of AI frameworks. This in turn entails a very high cost for companies in relation to the benefit provided by these technologies. And that is why quite a few firms have abandoned projects in the field of AI whose benefit did not justify the huge investment. In short, many companies have hired a multitude of engineers and spent huge amounts of money in resources as well as developing their own systems from scratch without having a solid Big Data strategy or ROI forecasts.
AI and its cognitive services are merely tools in a business toolbox. Tools that extract insights, find complex patterns, or optimize processes. We cannot pretend to solve all cases using these technologies alone without having a knowledge of the data and the business. Neither try to solve isolated problems that are part of a much larger digital entity nor continue to use analog information in a digital world. The potential of these high techs is still far from being fully exploited. Companies still offer a resistance to the massive and systematic use of AI technologies where they can play a paramount role when it comes to optimizing processes, extracting insights, and finding patterns in a continuous fashion (AI Augmented Analytics [25]). The new digital revolution will come from these AI technologies when they are used massively in our daily lives.
But how can I accelerate innovation and digital transformation within companies without the problems mentioned above? The answer lies in the use of AI cognitive services [26] in the cloud that are in continuous development, ready to be used as tools, require minimal effort to get up and running and can be easily reused in similar cases. AI cognitive services in the cloud allow having the most advanced AI technology at the service of digitalization and optimization of firms. They permit industrializing this type of AI for the entire company, enabling a deep digital transformation at all scales. Eliminates resistances and non-optimal processes. They automate and simplify procedures and data processing. But mostly perform repetitive and boring human cognitive tasks. The extensive use of these technologies in all areas will make companies more efficient and consequently reduce their expenditures. If there is any process that can be optimized or automated, companies will have no choice but to use these technologies if they want to have a full digital transformation and be a data driven company.
(Source Azure: https://azure.microsoft.com/en-us/products/cognitive-services/ )
In the case of Azure Cognitive Services [27], we can currently distinguish five main categories of services. Those that have to do with speech, language, and vision and making intelligent decisions as well as an additional category currently in preview that enables business to use the latest generation of large-scale AI models. These services use AI models pre-trained with Big Data that in turn can be customized for our particular use by adding our own data, all with minimal effort. Besides, AI cognitive services can be combined with each other or with other cloud services to develop solutions that fully adapt to our needs. And most importantly, they permit the reuse of systems in other similar processes as well as the training of the same data source for other use cases.
The AI cognitive services currently available in Azure by category are:
As previously mentioned, these services are in continuous development and expansion. (For more up-to-date information one can consult the official documentation [28])
One of our most recent cases of the use of AI cognitive services, which was quite a success story, has to do with automatic reporting of the use of construction machines in the railway sector. The client stored video recordings from all its surveillance cameras at more than 200 different sites. Video footage was utilized merely for surveillance and infrastructure security. Thanks to storage technologies and computer vision AI cognitive services in the Azure cloud, we were able to implement an automated system that would detect not only the type of machines and the hours of operation, but we could even report the model of the machine and all with an accuracy greater than 95%. The system avoids having to report the use of construction machines manually, which makes it more efficient since it has more precision and fewer errors than humans. The same data source was later reused to warn of human presence in dangerous areas with minimal extra development effort.
In other of our more recent cases, we used these cognitive AI technologies to be able to automatically classify legal documents. For another company, we were able to carry out a proof of concept that showed that the identification and retrieval of hand-filled entities in carbon-copy forms was possible. Even if the models had a very high accuracy, it was recommended to avoid whenever possible the use of analog information in our proposed solutions.
Swisscom Data & Analytics helps business customers with advisory, design, integration, and maintenance of analytical information systems such as data lakes, data warehouses, dashboards, reporting and ML/AI solutions based on selected technology from Microsoft, AWS, SAP, open source and more. More than 50 engaged data & analytics experts support our clients in different industries on a day-to-day basis in order to make them true data driven businesses.
Sergio Jimenez is a Senior Data & Analytics Consultant at Swisscom specialized in Advanced Analytics. Since joining Swisscom in 2016, Sergio has worked on numerous projects for multiple customers ranging from Business Intelligence to AI/ML. He has successfully built innovative solutions using the latest technologies.
[1] Big Data Analytics. IBM. Accessed Sep 2022. https://www.ibm.com/analytics/big-data-analytics
[2] Artificial Intelligence. IBM. Accessed Sep 2022. https://www.ibm.com/design/ai/basics/ai/
[3] Machine learning. IBM. Accessed Sep 2022. https://www.ibm.com/design/ai/basics/ml
[4] What is data lake. Microsoft. Accessed Sep 2022. https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-is-a-data-lake/
[5] https://hbr.org/2022/11/how-generative-ai-is-changing-creative-work
[6] https://research.ibm.com/interactive/generative-models/
[7] https://openai.com/dall-e-2/
[8] https://creator.nightcafe.studio/
[9] https://www.midjourney.com/
[10] https://beincrypto.com/learn/ai-image-generators/#h-1-midjourney
[11] https://deepai.org/machine-learning-model/text2img
[13] https://www.ibm.com/cloud/learn/conversational-ai
[14] https://aws.amazon.com/what-is/machine-translation/
[15] https://www.sciencedirect.com/science/article/pii/S1110866515000341
[16] https://www.ibm.com/cloud/learn/natural-language-processing
[17] https://www.ibm.com/topics/computer-vision
[19] https://techvera.com/6-ways-artificial-intelligence-can-cut-business-costs/
[20] https://techmonitor.ai/technology/ai-and-automation/call-centre-ai
[21] https://venturebeat.com/ai/why-do-87-of-data-science-projects-never-make-it-into-production/
[22] https://odsc.medium.com/machine-learning-challenges-you-might-not-see-coming-9e3ed893491f
[24] https://www.capgemini.com/gb-en/wp-content/uploads/sites/3/2017/09/dti-ai-report_final1-1.pdf
[25] https://powerbi.microsoft.com/en-us/augmented-analytics/
[26] https://digital6.tech/artificial-intelligence-ai-cognitive-services
[27] https://azure.microsoft.com/en-us/products/cognitive-services/
[28] https://azure.microsoft.com/en-us/products/cognitive-services/openai-service/#features
Sergio Jimenez-Otero
Senior Data & Analytics Consultant
Trouve le Job ou l’univers professionnel qui te convient. Où tu veux co-créer et évoluer.
Ce qui nous définit, c’est toi.