We understand the challenges our customers face within and across these industries. We provide practical, pragmatic and powerful solutions for addressing those challenges. These case studies show those solutions in action.
The importance of AI is recognized by NXP. NXP’s innovation team comprised intelligent people that had already been spending significant time on the topic, but none of the team had themselves a background in AI. Datamaister provided them knowledge to give a broader spectrum of AI techniques, methods and applications.
Datamaister provided a series of workshops, presentations, brainstorm sessions and benchmark tests for NXP.
NXP received feedback on what they had done so far, learned of new avenues for exploration and ideas on new implementations.
Datamaister’s broad knowledge of AI and ML can help a team to expand and improve their knowledge, methods and projects.
In an R&D setting: using bacterial DNA in soil samples as biomarkers to predict the presence of oil and gas deep beneath the earth.
Due to the imbalance between the number of samples (few) and number of features (many) we needed to define a combination of feature selection and deep learning. Feature selection was used to reduce to computational complexity and several other methods to prevent overfitting on the data.
A combination of feature selection, several deep learning models and stacking resulted in the most robust solutions.
Based on the available data, we delivered models which gave results between 80% and 90% accuracy. Furthermore we proposed an AI architecture which can easily manage new data and steadily improve its results.
Creating a test well is very expensive, therefore being able to reduce the number of wells saves a lot of money and effort.
Client: Large car manufacturer
Goal: Classify fault in paint on cars in factory
Tech: Tensorflow, Keras, feature creation
Classify faults in car coatings and to automatically label future faults.
Discovery workshop with experts. Simplify problem to make it solvable
(from 13 categories to 2)
Humans have an error of approximately 20%. Using votes from the experts to create labels. Create baseline solution. Improve solution using Deep Learning.
The solution of the deep learning methods was the best. Solution implemented the scanning robots. Robots give advice to employees
Reduced classification time and more time for correct labelling. The deep learning method has a 97% correct classification. 85% error reduction.
Cost savings and increased productivity because of the improved classification. Additionally the experts spent less time on the classification thus freeing up time for their other tasks.
Client: Large employment/recruitment agency
Industry: HR services industry
Goal:Automatically find the best candidates for a certain position
Tech: Solr, Lucene, NLTK, Scikit-learn
A job position is send to several HR companies. The client wanted to be the fastest to contact high potentials. Many candidates in their database we not recommended.
Discovery workshop to determine needs. Interactive creation of mockups. Robolect proposed solutions that the company had not thought possible.
We created a parser to extract the information from resumes. We implemented the feature creation mechanism and trained the system We integrated the solution to the automated sending of messages
One second after a job description arrives a ranked list of the best candidates is presented to the recruiter. The recruiter can decides what to do. The recruiters could focus on personal contact again.
The increased response time increased the number of vacancies filled, increasing their market share The recruiters can spend more time on their greatest added value: personal contact.
Artificial Intelligence can help you innovate your business case or solve critical business challenges through leveraging the power of machine learning. Combining your business knowledge with our AI expertise, we can find the best way to solve your critical business challenges.
Feel free to contact us for a no-obligation conversation in which we will be happy to inform you further about our consultancy services in the field
of Artificial Intelligence and Data Science.