Clément Stenac is an avid software engineer, Chief Technology Officer and co-founder of Dataiku. He oversaw the design and development of Dataiku, software that uses data and AI everyday behavior for everyone. Clément was previously at the forefront of Exalead’s product development, leading the design and implementation of web-scale search engine software. He also has extensive experience in open source software, as a former developer of VideoLAN (VLC) and Debian projects.
Computer vision is a powerful data science and machine learning area that uses deep learning models to understand the content of images and video.
Instructing a computer to recognize and analyze the content of images can streamline manual processes and reduce time to make decisions, as well as open up powerful applications and new ones. way of interacting with customers. These cutting-edge techniques are not available to many companies, but with recent innovations, many organizations and users are able to drive value using computer vision techniques.
Traditionally, deep learning, including computer vision models, has been the domain of expert data scientists who use advanced frameworks such as TensorFlow or PyTorch and custom code to create models. The usual nature of these models means that it can take months to develop and can be challenging to update and maintain, often taking weeks for the simplest changes.
Training computer vision models also requires annotated images that show the various objects or conditions on which the system will learn. People prepare these training images by reviewing each image and then identifying the relevant information. This process can be time consuming for data science teams to find experts, provide images and annotation instructions, and then review and format the relevant information. These challenges prevent many companies from taking advantage of computer vision techniques, which limit how they can improve processes and customer experiences.
Insurance Case Study
One example where in-depth learning, including computer vision and natural language processing techniques, is being successfully used today is the review of insurance claims. Traditional auto insurance claims go through a manual review that can take days or weeks to determine the appropriate action. For consumers who have recently been in a traumatic car accident, delays in processing a claim can only exacerbate things and can cause customer frustration and create negative sentiment in the company. .
A well-known insurance company is streamlining the processing of auto claims to provide near-real-time responses using Dataiku. The insurance provider previously used days or even weeks to process claims. But now, when customers call about claims, the phone claims agent can then immediately tell them if their claim will result in a total loss or repair of their car. and take them to the next step immediately.
Instructing a computer to recognize and analyze the content of images can streamline manual processes and reduce time to make decisions, as well as open up powerful applications and new ones. way of interacting with customers.
Machine learning models help run this almost instantaneous background process by reading and processing acquisition documents and reviewing any images of the damaged vehicle. The combination of structured information provided by the customer on the claim form, unstructured text describing the accident and condition of the vehicle, and pictures of the vehicle is sufficient for the machine learning model to determine if the damage is too severe to repair.
This change in processing time converts the people who handle claims from form fillers who navigate customers through a complicated and sometimes painful claims process into customer heroes who can help them immediately and provide concrete next steps.
New Computer Vision Capabilities Enable New Users
Dataiku is a data science and machine learning platform used by companies around the world to create, deploy, and manage machine learning and deep learning models. Dataiku is known for enabling a variety of technical and non-technical users to take on machine learning projects with a collaborative environment that supports everything from full code to zero. user code.
Announced in the recent release of Dataiku 11 are new computer vision modeling capabilities that allow non-technical users to create models using a visual, no-code interface. Users can configure the image training system, and the AutoML engine takes care of the rest. To help create training images, Dataiku also introduced a new managed labeling system that allows data science teams or program administrators to assign labeling tasks to groups of subject matter experts and monitor the development and quality of the label.
Users can use computer vision modeling in a variety of use cases in industries where real-time image processing can save time and money. For example, computer vision models can improve manufacturing quality control, speeding up the processing of everything from computer chips to jewelry. This helps to easily recycle defective items and reduce issues for customers. Construction companies use computer vision to monitor site safety, comply with OSHA regulations, and limit downtime and equipment loss by monitoring equipment and ensuring workers wear safety equipment. .
Computer vision used to be the domain of experts, but projects can take a long time to develop and challenge sustainability. However, with the changes in the technology landscape, computer vision is fast becoming an area that many companies can take advantage of to provide better customer experiences and shorter cost.
New Stack is a wholly owned subsidiary of Insight Partners, an investor in the following companies mentioned in this article: Enable, Dataiku.