Categoría: Artificial Intelligence

On April 25th I was invited to join the “Seminar on Artificial Intelligence Applied Technologies in the Insurance Industry”, together with two speakers: Adel Abed, from FID Seguros and Héctor Monje, from Microsoft.

This event was organized by InsurteChile and the Pontificia Universidad Católica de Chile. Its purpose was to explore Artificial Intelligence, delving into the benefits and applications within the insurance industry.

Although you might think that Artificial Intelligence is a recent concept, the truth is that it dates back to 1956 and is based on analytical techniques and the combination of algorithms in machines that can have the ability to resemble human intelligence.

It should be noted that while AI has been around for a long time, the mathematical models did not have the same scope as they do today due to the availability of data. Also, previously this technology did not have such a «sexy» name as it does now: Artificial Intelligence.

This progressive evolution of technology has led insurance industries to evolve from being robust and complex (like a medieval warrior), to being agile and accurate.

The digital era was a key element for insurers, one of the most traditional and conservative industries in existence, to decide to transform themselves with the help of insurtechs. This with the aim of improving their processes and, ultimately, offering a unique service to their policyholders.

Uses of Artificial Intelligence in the Insurance Industry

Today it is possible to see the application of artificial intelligence in practically the entire value chain of insurers:

• Price optimization.
• Competitor pricing.
• Churn (Cox survival).
• Life time value for customers and brokers.
• Conversion and renewal probability.
• Automatic claims provisions.
• Data governance and unique reporting.
• Segmentation of clients and brokers.
• Xsell and Upsell probability.
• Chatbots and SpeechBots for self-care and service.
• Fraud prediction and detection (text and image analysis in inspections, social networks).
• Parameterized insurance (gives way to indemnity by weather thresholds).

What is the impact of AI on LISA Insurtech?

LISA Insurtech was born precisely in the search to help insurance carriers accelerate their claims processes in an automated way with Artificial Intelligence. Through  LISA Claims we are in charge of catalyzing the efficiency of insurance processes that are usually robust and complex. The idea is to streamline them, reduce costs and ensure that the customer, at the moment of truth (at the time of enforcing their policy after a claim), is really satisfied.

We achieve this through text and image recognition combined with a mathematical model to make the insurance process efficient. A more agile, faster and lighter process.

How do we operate?

The documentation that is entered into our core feeds our database at all times and our Artificial Intelligence with Natural Language Processing (NLP) is present there: 

• We identify the type of documentation.
• We recognize what data you need.
• We distinguish what you need to retrieve.
• We feed the database (helps us to make future decisions).

Our AI trainings are scalable and allow us to provide a timely response to our customers.

What do you achieve with the application of AI?

• Ability to make agile and automatic decisions.
• Efficient use of resources and better data quality.
• Certainty in quick responses and timely attention.

The AI and the challenges behind

• How digitized are companies? This will make it more or less difficult to follow through with the transformation.
• Understanding and managing expectations: Step by step so that the implementation can be cemented and, over time, transformed.
• Conviction: Ask yourself: Are you really ready to transform the industry?

Before ending this article, i would like to thank InsurteChile and the Pontificia Universidad Católica de Chile for being generators of this type of spaces so important for the transformation of the insurance industry. I would also like to express my pleasure in joining Adel Abed from FID Seguros and Héctor Monje from Microsoft, who shared with us very important information on the subject.

Today we share with you 6 other technology trends that are going to be very relevant for this coming 2022! As we mentioned in the previous article, technology is not only here to stay, but also to transform itself every day to be its best version and positively impact many companies.

Now lets start!

6. Software-defined networks

These are a set of techniques related to the area of computer networks, whose objective is to facilitate the implementation and deployment of network services in a dynamic and scalable way. The aim is to avoid the network administrator having to manage these services at a low level.

According to the vmware article, software-defined networking (SDN) represents an approach in which networks use software-based controllers, or application programming interfaces (APIs), to direct network traffic and communicate with the underlying hardware infrastructure.

OpenFlow and SDN will make networks more secure, transparent, flexible and functional.

7. The Cloud

According to the Eninetworks article, «What is the cloud and how is it used?«, it is a data storage service to servers located on the network. This allows programs and files to be uploaded, opened, modified or used through a connection without the need for them to be located on the storage of the device being used.

One of the key issues for the insurance industry is to transform insurance companies into data-driven organizations. Cloud computing adoption is increasingly being driven by the need for digital transformation and being perceived as a business enabler due to its large capacity to store information.

By 2022, the cloud will be more entrenched and more novel. However, the possibility of increasing capabilities and working in the cloud is still unlimited. 

8. Internet of things

The Internet of Things, or IoT for short, is a key factor and part of insurers’ transformation to the Digital Age, as the study and analysis of this field is what will define what is important to people.

It is a technology that connects to the Internet our household objects, those that are of daily use and that are ordinary. However, this connection to the Internet is not enough; the IoT is about connecting objects to each other.

This is how different objects will be able to interact with each other as part of a program. For example, the alarm clock can be synchronized with the coffee maker, and once the alarm is activated, the coffee maker can simultaneously start brewing coffee.

In this way, the relationship with our environment will change drastically. Thus, our clothes could give us an analysis of our biometric data or our homes could carry out the purchase of basic products in an automated way. In addition, the connection between the elements of our environment would ensure greater individual security. 

9. Big Data 

This term describes the large volume of data, both structured and unstructured, that occupies large parts of the business. What really matters about big data is what companies can do with the data, because by analyzing it they can gain insights that lead to better decisions and strategic business moves.

10. Machine learning 

Machine Learning is one of the applications of AI that provides systems with the ability to learn and automatically improve the experience without having to be programmed. Thanks to this technology we can plan company resources, develop predictive systems on customer behavior and personalize communications, generating greater value for users. 

Machine learning plays an increasingly important role in our lives, whether to classify search results, recommend products or create better models of the environment.

11. Computer vision and pattern recognition

As the viso.ai article explains, it is a field of Artificial Intelligence (AI), which deals with computational methods to help computers understand and interpret the content of digital images. 

Thus, Computer vision aims to make computers see and understand visual data input from cameras or video sensors. This is to help computers automatically understand the visual world by simulating human vision using computational methods.

Finally…

It all boils down to the ability of companies to keep pace with technological innovations, to adopt them and take advantage of them.

Technology is going by leaps and bounds and does not stop, which is why it is so important to know it and take advantage of it. This is not limited to a single industry sector, but can even be taken to insurance companies.

How so?

It is no secret that the insurance industry is one of the oldest and most difficult industries to add technology to its processes. That is why LISA Claims was born with the aim of streamlining those processes and offering a better service to its policyholders.

LISA Claims is a platform that controls and manages all claims settlement processes through the use of technology, guaranteeing security, improved operational efficiency and greater policyholder satisfaction.

Thanks to technology, it is easy to innovate and transform tasks that previously depended on human labor into automated tasks with the help and fusion of artificial intelligence and an automaton.

What technologies does LISA Claims use?

• Computer security.
• Big data.
• Cloud Computing.
• Machine learning.
• Computer vision.
• Internet of things.
• Interconnections and digital ecosystems.

As we explained in our previous article, object detection is one of the star technologies of Artificial Intelligence. Now, in this second part, we would like to show you the use cases of object detection and its influence on LISA Insurtech.

Object detection use cases + important

The use cases involving object detection are very diverse.

Did you know that this technology has been implemented in computer vision programs for various applications, from sports production to productivity analysis?

Today, object recognition is at the core of most vision-based artificial intelligence programs and software. In addition, object detection plays a vital role in scene understanding, which is popular in security, transportation, medical, and other use cases.

According to the article «Object Detection in 2021: The Definitive Guide» some of the most relevant use cases are:

• Autonomous driving: Autonomous vehicles rely on object detection to recognize pedestrians, traffic signs, other vehicles and more. For example, Tesla’s Autopilot AI makes extensive use of object detection to sense approaching vehicles or obstacles.
• Vehicle detection with AI in transportation: Object recognition is used to detect and count vehicles for traffic analysis or to detect cars stopping in dangerous areas, e.g., at intersections or roadways.
• Medical feature detection in healthcare: Medical diagnoses rely heavily on the study of images, scans and photographs. In the same vein, object detection involving CT and MRI scans has become extremely useful for diagnosing diseases, for example, with ML algorithms for tumor detection.

Finally….

Object detection is one of computer vision’s most fundamental and challenging technologies. It has received significant attention in recent years, especially with the success of deep learning methods that now dominate the latest detection methods.

Object detection is becoming increasingly important for computer vision applications in any industry.

Before concluding, we would like to open an essential point since object detection within the insurance industry could offer competitive services, for example, by detecting cars and evaluating their driving behavior (and seeing why there are more or fewer traffic accidents).

It also serves as a support when evaluating an auto or home claim, but how?

At LISA Insurtech, we specialize in streamlining all claims processes with cutting-edge technology, which is why we rely on artificial intelligence, machine learning, and deep learning. From this, we can analyze photographic and video evidence to prevent fraud and accurately calculate the damage’s value.

Would you like to learn more about our flagship product LISA Claims?

Object detection is a critical field in Artificial Intelligence (AI), as it allows computer systems to «see» and «comprehend» their environments by detecting objects in visual images or videos.

It is used to detect visual objects of various kinds (humans, animals, cars, or buildings) in digital images such as photos or video frames. Its goal is to develop computational models that provide the most basic information computer vision applications need (where the objects are and what they are doing).

Why is object detection important?

Object detection is one of the cornerstones of computer vision and forms the basis for many other subsequent tasks—for instance, segmentation, image captioning, object tracking, and more. 

Specific object detection applications include pedestrian detection, people counting, and face and text detection. 

Object detection + deep learning

Rapid advances in deep learning techniques have greatly accelerated the momentum of object detection. With deep learning networks and the computing power of GPUs, the performance of object detectors and trackers has improved tremendously, achieving significant advances in object detection.

Machine learning (ML) is a branch of artificial intelligence (AI). It involves learning patterns from examples or sample data as the machine accesses these and can learn from them (supervised learning on annotated images).

How does object detection work?

Thanks to an article from viso.ai, we can learn that object detection can be performed using traditional image processing techniques or modern deep learning networks:

1. Traditional image processing: Image processing techniques generally do not require historical data for training and are unsupervised in nature.

• Advantages: These tasks do not require annotated images, where humans manually label the data (for supervised training).
• Disadvantages: These techniques are restricted to multiple factors, such as complex scenarios (no single-color background), occlusion (partially hidden objects), illumination and shadows, and clutter effect.

2. Modern deep learning networks: Deep learning methods rely on supervised training. Performance is limited by the computational power of GPUs, which is rapidly increasing year after year.

• Advantages: Deep learning object detection is significantly more resilient to occlusion, complex scenes, and challenging lighting.
• Disadvantages: A large amount of training data is required (plus the image annotation process is laborious and expensive).

Finally…

Object detection is one of computer vision’s most fundamental and challenging technologies. Consequently, it has received significant attention in recent years, especially with the success of deep learning methods that now dominate the latest detection methods.

Object detection is becoming increasingly important for computer vision applications in any industry.

Stay tuned! Our next article will tell you about the use cases and how we employ them at LISA Insurtech.

You can’t miss it!

What is Edge Intelligence and Edge AI?

The combination of Edge Computing and AI has given rise to a new area of research called «Edge Intelligence» or «Edge AI». Edge Intelligence makes use of pervasive edge resources to power artificial intelligence applications without relying entirely on the cloud.

While the term Edge AI or Edge Intelligence is completely new, practices have started early.

However, despite the early start of exploration, there is still no formal definition of edge intelligence.

Currently, most organizations and printers refer to Edge Intelligence as «the paradigm of running AI algorithms locally on an end device, with data (sensor or signal data) being created on the device.»

Here’s a curious fact…

Several major companies and technology leaders, including Google, Microsoft, IBM and Intel, demonstrated the benefits of edge computing. Their efforts include a wide range of artificial intelligence applications:

• Real-time video analysis
• Cognitive assistance
• Precision agriculture
• Smart home
• Industrial IoT.

Advantages of pushing deep learning to the edge

Thanks to the viso.ai article we were able to learn that the merger of artificial intelligence and edge computing is natural, as there is a clear intersection between them. Data generated at the edge of the network relies on artificial intelligence to fully unlock its potential and edge computing can thrive with richer application and data scenarios.

The advantages of implementing deep learning at the perimeter include:

1. Low latency: Deep learning services are deployed close to the requesting users. This significantly reduces latency and the cost of sending data to the cloud for processing.
2. Privacy preservation: privacy is enhanced as the raw data needed for deep learning services is stored locally rather than in the cloud.
3. Increased reliability: decentralized, hierarchical computing architecture provides more reliable deep learning computing.
4. Scalable deep learning: with richer application and data scenarios, edge computing can promote the application of deep learning and drive the adoption of artificial intelligence.
5. Commercialization: diversified and valuable deep learning services extend the commercial value of edge computing and accelerate its deployment and growth.

Finally…

Technology has always advanced by leaps and bounds. With the emergence of both Artificial Intelligence and IoT, the need arises to push the frontier of the former from the cloud to the Edge Computing device.

Edge computing has been a widely recognized solution to support compute-intensive machine vision and artificial intelligence applications in resource-constrained environments.

LISA Insurtech

We believe that keeping up with technology will not only provide us with many benefits. This favors the top of mind of traditional sectors such as insurance, which has been one of the last to acquire technology.

Therefore, our promise has always been to pursue the technology of insurers with artificial intelligence, machine learning, deep learning, big data, the cloud, among others.

With the advances in technology, we have witnessed in recent years a boom in artificial intelligence (AI) applications and services such as edge computing. All this thanks to the momentum of advances in mobile computing and the Internet of Things (IoT), where billions of mobile and IoT devices are connected to the Internet, generating trillions of bytes of data.

Undoubtedly, there is an urgent need to bring AI to the edge of the network to fully unlock the potential of Big Data at the edge. To realize this trend, Edge Computing is a promising solution to support compute-intensive AI applications on edge devices.

Edge Intelligence or Edge AI is a combination of AI and Edge Computing, which allows the implementation of machine learning algorithms in the edge device where the data is generated.

What is Edge Computing?

Edge computing is the concept of capturing, storing, processing and analyzing data closer to the location where it is needed, in order to improve response times and save bandwidth.

In that sense, edge computing is a distributed computing framework that brings applications closer to data sources, such as IoT devices, local end devices or edge servers.

As noted in NetworkWorld, Edge Computing «allows data produced by Internet of Things devices to be processed closer to where it was created rather than being sent over long distances to reach data centers and compute clouds.»

Why do we need Edge Computing?

As a key driver pushing the development of AI, Big data has recently undergone a radical shift of data sources from large-scale cloud data centers to increasingly pervasive end devices such as mobile and IoT devices.

Traditionally, big data, such as online shopping records, social media content and enterprise computing, was born and stored primarily in large-scale data centers. However, with the emergence of mobile computing and IoT, the trend is now reversing.

Today, a plethora of sensors and smart devices generate massive amounts of data, and ever-increasing computing power is driving core computation and services from the cloud to the network edge. 

Did you know?

Today, more than 50 billion IoT devices are connected to the Internet, and it is predicted that by 2025, 80 billion IoT devices and sensors will be online.

What is edge intelligence and what are its advantages? Don’t miss the topic in our next article – expect it this week!

Image recognition technology with AI is becoming more and more indispensable in any company you can imagine. Its applications bring economic value in sectors such as healthcare, retail, security, agriculture and many others..

In this article you will learn more about Image recognition use cases.

Face identification and analysis

It is a leading image recognition application. Modern ML methods make it possible to use the video feed from any digital camera or webcam..

In these applications, image recognition software employs AI algorithms for simultaneous face detection, face pose estimation, gender and age recognition using a deep convolutional neural network.

Facial analysis with computer vision enables systems to recognize identity, intentions, emotional and health states, age or ethnicity. Some photo recognition tools seek to quantify perceived attractiveness levels with a score.

Medical image analysis

Visual recognition technology is widely used in the medical industry to enable computers to understand images that are routinely acquired throughout the course of treatment.

For example, there are multiple papers on the identification of melanoma, a deadly skin cancer. Deep learning image recognition software enables tumor tracking over time.

Animal monitoring

Agricultural visual AI systems use novel techniques that have been trained to detect the type of animal and its actions. AI image recognition software used for remote monitoring of animals in agriculture for the detection of diseases, anomalies, compliance with animal welfare guidelines, etc.

Pattern and object detection

AI photo and video recognition technologies are useful for identifying people, patterns, logos, objects, places, colors and shapes. The customizability of image recognition allows it to be used in conjunction with multiple software programs.. 

For example, after an image recognition program specializes in people detection, it can be used for people counting, a popular computer vision application in retail stores.

Automated plant image identification

Thanks to an article from viso.ai, we can learn that in a July 2021 research analyzed the accuracy of image identification. This in order to determine plant family, growth forms, life forms, etc.

The tool works by using the photo of a plant with an image comparison software to query the results with an online database.

The results indicated a high recognition accuracy, as 79.6% of the 542 species from about 1,500 photos were correctly identified. While the plant family was correctly identified for 95% of the species.

Image recognition is very favorable and relevant for any company since it speeds up many processes, favors the collection of data and also the work without human hands.

Making way for technology and all that image recognition entails is not an easy task, but with knowledge, a good organization and a specialized team, it will be possible.

At LISA Insurtech we stand out for streamlining insurance settlement processes with cutting-edge technology. One of our star performers is our Artificial Intelligence.l.Thanks to it, we are able to recognize images, documents, videos and photographs in order to avoid fraud and avoid so many frictions during the claim settlement.

Image recognition with deep learning is a key application of AI vision and is used to drive a wide range of real-world use cases today.

Image recognition with deep learning is a key application of AI vision and is used to drive a wide range of real-world use cases today.

What is image recognition?

Simply put, it is the task of identifying objects of interest within an image and recognizing to which category they belong. Photo recognition and image recognition are terms that are used interchangeably.

When we visually detect an object or scene, we automatically identify objects as distinct instances and associate them with individual definitions. However, visual recognition is a very complex task for machines.

​​Image recognition using artificial intelligence is a long-standing research topic in the field of computer vision. Although different methods have evolved over time, the common goal of image recognition is the classification of detected objects into different categories (also referred to as object recognition).

In recent years, machine learning, in particular deep learning technology, achieved great successes in many computer vision and image understanding tasks.

What is image recognition used for?

Across all industries, AI image recognition technology is becoming increasingly indispensable. Its applications bring economic value in sectors such as healthcare, retail, security, agriculture and many more.

Three most popular image recognition machine learning models

Thanks to the viso.ai article we were able to learn about these three most popular types of models:

Support vector machines

SVMs work by making histograms from images that contain the target objects and also from images that do not. The algorithm then takes the test image and compares the trained histogram values with those of various parts of the image to check for matches.

Feature bag models

These models, such as scale invariant feature transform (SIFT) and maximally stable extreme regions (MSER), work by taking as a reference the image to be scanned and a sample photo of the object to be found. It then attempts to match features in the sample photo to various parts of the target image to see if matches are found.

Viola-Jones Algorithm

A face recognition algorithm widely used in the era before convolutional neural networks, it works by scanning faces and extracting features that are then passed through a boosting classifier. This, in turn, generates a series of boosted classifiers that are used to check test images.

To find a successful match, a test image must generate a positive result from each of these classifiers.

Deep learning image recognition models

The most popular deep learning models, such as YOLO, SSD and RCNN, use convolution layers to analyze an image or photograph. During training, each convolution layer acts as a filter that learns to recognize some aspect of the image before moving on to the next.

One layer processes the colors, another the shapes, and so on. At the end, a composite result of all these layers is taken into account to determine if a match has been found.

This topic is quite extensive! But we have made for you a series of articles with compressed information that will teach you everything you need to know about image recognition.

Stay tuned and don’t miss it!

In the previous article, we were already able to know the stages of evolution of computer vision over the years. Now, we want to explain in this opportunity the 6 use cases of this technology.

There are many companies that are thinking or have been rapidly introducing machine vision technology in various sectors to solve automation problems with computers that have the ability to observe. 

That is why we want to show you in this article, which are the sectors that mostly implement this technology and for this we rely on the viso.ai article.

Manufacturing:

Computer vision is used in manufacturing industries for automated product inspection, object counting, process automation and to increase worker safety.

Security:

People detection is performed for intelligent perimeter surveillance. Another popular use case is deep face detection and facial recognition with better-than-human accuracy.

Agriculture:

This sector includes automated animal monitoring to detect animal welfare and early detection of diseases and abnormalities.

Smart cities:

It is applied to crowd analysis, weapons detection, traffic analysis and vehicle counting, and infrastructure inspection.

Retail trade:

For example, video from retail store surveillance cameras can be used to track customer movement patterns and for people counting. It can also be used to analyze footfall to identify bottlenecks, customer service and waiting times.

Insurance:

We open this point especially because we know for sure that machine vision can be used in the insurance industry.

How? It can be applied to speed up settlement processes by analyzing photographic, documentary and video evidence submitted by policyholders. It also allows fraud prevention.

Currently, artificial vision is in an enviable state of progress and possibilities that has allowed us to reach more complex and robust applications such as LISA Claims.

Through our platform, we can control and manage all claims settlement processes thanks to the use of technology, guaranteeing:

Security, improved operational efficiency and greater policyholder satisfaction.

Computer vision has come a long way since its inception in the 1960s when computer scientists first attempted to emulate human sight through computation. In recent years, new deep learning technologies have led to significant advances in computer vision, particularly in image recognition and object detection.

Let’s take a closer look at the evolution of this technology:

1960s: Computer vision research began, but recognizing multiple natural objects with variations in shape was still a challenge.

2014: Deep learning technology and training computers with millions of images from the largest image classification dataset (ImageNet) led to breakthroughs. Deep learning demonstrated superiority over traditional algorithms in tests and challenges.

2016: Deep learning became faster and more efficient with the development of single-stage object detectors, using multilayer convolutional neural networks (CNNs) to simplify feature extraction and description. CNNs became the de facto standard computing framework for computer vision.

2020: Deep learning and edge AI were deployed, enabling computer vision to be realized on low-cost hardware and mobile devices. Numerous deeper and more complex networks were developed for CNNs to achieve near-human accuracy in many computer vision applications.

Computer vision has come a long way, from recognizing common objects to almost human vision. These kind of technologies have evolved significantly with extensive testing, studies, and data.

In our next article, we will showcase six real-life cases of computer vision use in different industries. Don’t miss out!

If you’re interested in learning more about artificial intelligence and image recognition and its role in the future of technology, make sure to follow our blog for the latest news and insights.