Artificial Intelligence and Machine Learning is no child’s play. But who said that it’s not fun.
We explore new possibilities and go beyond to seek answers to things that aren’t visible. All with a childlike wonder and with the curiosity of a teenager. We play our part to turn ideas into reality, to solve business challenges, and to develop MVPs. Some call it a workplace, we call it our playground.
This is where we make things happen.
Analyzed various deep learning networks to identify the most suitable that could detect plastic from paper. Results from just a cell-phone capture allowed the customer to move away from spectral imaging and CV and consider deep learning networks as a credible alternative. A technical paper explaining the Neural Network model and the implementation details can be found here.
A US based business with manufacturing facility in Mexico wanted a system that could count the number of people within a given facility at any given point in time. People could be sitting, standing, moving and may be only partially visible. Our system delivered 100% accuracy for their facility.
An automated system that is capable of questioning participants in a clinical/drug trial and extracting intelligence from responses that otherwise would have been missed out by traditional systems that use an app or questions with multiple choice answers.
A learning-system for biomechanical determination of correctness of exercise posture and range of motion. Eventually to be used to monitor/evaluate participants in a single or a group exercise class. Reduces the need for multiple instructors and allows each participant to monitor their own progress in detail.
Demonstrates use of voice commands for operating equipment through the amazon ecosystem (Alexa). The demo system was rapidly developed harnessing the simplicity of our proprietary WhirlBot engine (an omni-channel chatbot)
The Innovations team of a major US based clothing retailer approached us to build a proof of concept to provide the online shopping experience in a physical store with an AR approach. A deep learning neural network was built to identify a product through the customer’s phone camera and display the product information from the retailer’s website.
Demonstrates use of voice commands for browsing and shopping on an ecommerce website. The demo system was developed harnessing the simplicity of our proprietary WhirlBot engine (an omni-channel chatbot).