Your AI Challenge
is our Business

This project involved the development of an image processing tool to detect design and production flaws in industrial Printed Circuit Boards (PCBs). The tool uses up sampling, or super resolution, techniques to improve the resolution and detail of the images, allowing for more accurate inspection. By identifying and correcting mistakes during the production process, this tool helps to save the client a significant amount of money.

This project involved the development of a computer vision neural network for use in video conferencing applications. The AI model was designed to provide cognitive insights and visibility into various aspects of the meeting, such as deals, team performance, and market changes. By analyzing the visual and auditory data from the video conference, the neural network is able to extract relevant information and provide insights that can be useful for decision-making and improving team productivity.

This project involved the development of a training pipeline to optimize the control of autonomous vehicles for low energy consumption. The pipeline includes machine learning algorithms and techniques that are specifically designed to reduce energy usage while maintaining high performance and safety to extend their range and reduce the overall cost of operation.

This project involved the development of AI and data science techniques for use in the mobile marketing platform, AppsFlyer. The focus of the project was on analyzing commercial cartoons, with the goal of increasing engagement with marketing materials. This involved the use of machine learning algorithms and data science techniques to identify trends and patterns in the data, and to develop insights into what factors contribute to successful cartoon marketing campaigns. 

The project for 1etx involves the development of a full stack labeling tool and training process for analyzing chemical interactions in electrophoresis. The tool includes a convolutional neural network for detecting DNA cleavage. The training process uses the labeling tool to annotate and classify chemical interactions, allowing the neural network to learn and accurately identify DNA cleavage. This improves the accuracy and efficiency of DNA analysis in research and clinical settings.

The project involves the development of a robotics device for capturing and analyzing high resolution color images of the fundus, the inner lining of the eye. The device uses near infrared light to aim the camera and find the ideal position for image capture. These images can be used for the early detection of retinopathy, a group of disorders that can cause vision loss. Automating the process of fundus imaging improves accuracy and efficiency of retinopathy detection for earlier treatment and better patient outcomes.

This project involved the provision of full time R&D services for the creation of AI models for anomaly detection using street cameras. The goal of the project was to develop machine learning algorithms and models that could analyze video footage from street cameras in real-time, and identify unusual or suspicious behavior or events. These models could be used by security agencies, law enforcement, and other organizations to enhance public safety and security. The R&D services provided involved the full lifecycle of model development, from initial research and prototyping to final testing and deployment.

This project involved the development of real-time signal processing algorithms for chord detection on edge devices. The chord detection algorithms developed in this project are designed to analyze audio signals in real-time and identify the chords being played, allowing for applications such as music education, transcription, and analysis. The use of edge devices enables the chord detection to be performed efficiently and without the need for a constant internet connection.

The project involves the use of AI and data science techniques for analyzing urine samples with a low cost smart urinal spectrometer. The spectrometer uses light to detect markers for various health conditions. Machine learning algorithms and data science techniques are used to interpret the data and identify patterns relevant for diagnosis and treatment. The goal is to improve the accuracy and efficiency of urine analysis for better patient outcomes.

The project involves the development of medical biological data science techniques for cancer immunotherapies and target discovery. The goal is to predict patient responses to Immuno-Oncology treatments and identify new drug combinations for disrupting resistance in non-responding patients. The project analyzed medical and biological data, including data on cancer immunotherapies, patient characteristics, and treatment outcomes. Machine learning algorithms and data science techniques were used to identify trends and patterns in the data and optimize treatment strategies for improved patient outcomes.