One of our clients is Austria's biggest manufacturer and distributor of innovative, analytical instruments. Their instruments are offered to customers who want to monitor the performance of their manufacturing equipment. One of their products, EasyDens, is used for home brewers.
They came to us to make an AI product that would predict when maintenance would need to be performed. This approach delivered major cost savings over routine or other time-based, preventive maintenance.
Since implementing the AI PM product, the client has helped its clients improved maintenance planning. They worked their way out a reactive cycle into true preventative & predictive maintenance.
Tech Stack used Python, Autoencoder, One-Class SVM, Pyod, Multivariate Gaussian, TensorFlow, Keras.
The transformation resulted in a reduced maintenance cost of up to 25% for their clients.
One of our clients based in the USA is a cutting-edge, size recommendation and visualization platform for fashion eCommerce. They wanted to enhance the capability of their product to find the perfect fit for brand shoppers.
Their Computer Vision Body Measurement AI tool now helps their brands increase conversion and reduce returns. It increases time efficiency and improves accessibility to the process of trying on clothes. No longer just in the dressing room !
Since implementing the Computer Vision AI product, the client has helped its brands reduce returns and generate happy customers.
Tech Stack used Python, TensorFlow, Keras, Pytorch, fastai, openCV, CNN.
The transformation resulted in a reduction of sales returns for their brands of up to 20%..
One of our clients based in India is an eCommerce retailer. They want to be able to identify high-value customers, keen to buy on their platform, so as to optimize their journey and to optimize marketing resources, leading them to increased profits.
First, the important customers are identified. Then the mapping of these customers' interactions through different channels with the brand is identified. A value of importance Is assigned to each of the channels. Lastly, credit for the conversion is assigned to particular channels over others.
Tech Stack Python, Markov Chain
The transformation resulted in increased sales by 15% and it lowered the rate of cart-abandonment by 18% .
One of our Indian clients is a power distribution and support services company. Their service level went down dramatically due to an unforseen demand spike for electrical technicians.
That led them to want to be able to predict manpower requirements for their technicians and linemen, 30 days into the future. This would help them improve their service level while keeping cost under control.
We assisted them in building a product that could forecast their manpower requirement for the next days, using deep learning techniques.
Tech Stack Python, ARIMA, LSTM, RNN, Fuzzy Time Series, Vector Auto Regressive .
Their transformation improved their forecasting accuracy with up to 90%.
