"Third Opinion Platform"Third Opinion
is the developer of innovative solutions for the healthcare system. We create products based on machine learning technologies that improve the quality of medical services in the field of medical diagnostics and video analytics.
We are the leader of the Russian market with ambitions to enter the international market within the next two years. And due to the expansion of key areas, we are opening two new vacancies for specialists in the field of computer vision at the Middle level. What is to be done within the framework of two separate directions
- KT of the chest organs - segmentation, detection of COVID-19 signs and other pathological signs on 3D chest images;
- KT of the brain- construction of classification and segmentation models for determining stroke, hemorrhages and other pathologies on 3D brain images.
It will be a plus:
- CV expertise - 1-3 years of work experience;
- classic stack: numpy, scipy, pandas, cvv2, etc.;
- something from Pwtorch/MXNet / TensorFlow + DALI;
- knowledge of Python;
- experience working with segmentation models (Unet, DeepLab, etc.) and solving classification and detection problems;
- be able to use linux, git and docker;
- the ability to deal with SOTA approaches and quickly implement articles;
- support of the development process at all stages - from setting up experiments to integration into products.
- experience with medical data;
- experience with ONNX and TensorRT ;
- experience in writing scientific articles;
- practical experience in NLP and/or Speech Recognition.
- flexible schedule 5/2 with the possibility of choosing the beginning of the working day;
- office - Krasnopresnensky business center, 10 minutes from m. 1905 Street;
- medical expertise in the person of leading Russian doctors who carry out continuous data markup and consultations;
- workplace with gpu and own servers for training;
- the trial period is 3 months.
Send your SV and questions to the mail email@example.com
with the subject "CV from web" or write to telegram: @allgrom