Data Science Engineer

1. Design, build and maintain efficient, realiable and reusable codes.
2. Document funcitonal all aspects of the product design process.
3. Able to integrate with other platform (like C++, embedded devices, Cloud etc)
4. Responsible for ensuring the overall quality of technical deliverables / artifacts.
5. Communicating and managing work in a cross-funcitonal environment.
6. Ability to think out-of-the-box to achieve goals.
7. Strategise the logical interpretations before get involved in trainig environment.
8. Understanding by reading many research papers (e.g arxiv, CVPR, ICCV, ICML etc)
9. Opportunity to grasp and understanding for creating IP-standard product.
10. One step further to strategize the training process for less data or no-data.
11. Must have eager to learn and team-spirits

Skills & Qualifications

1. B.Tech / B.E/ Masters in Computer Science, engineering or related field
2. Strong Python programming, Linux OS and shell scripting skills.
3. HandsOn experience in Model Life Cycle management and accelerating the models on various hardware platforms like GPU/FPGA/TPU etc.
4. HandsOn experience with Docker, Kubernetes, GCP/AWS/Azure
5. Knowledge of deep learning architectures like CNN, RNN, LSTM, RL, GANs, Attention Models, Transfer Learning etc.
6. Solid grasp of neural architecture (e.g VGG/GoogleNet/MobileNet/SqueezeNet/EfficientNet etc) with pruning layers, modifying layers (with weights understanding) and quantization (by achieving accuracy).
7. Development skills with ML frameworks e.g PyTorch, Caffe, TensorFlow, Keras, DeepStream, scikit-learn etc.
8. Good understanding on model conversions (like caffemodel, onnx, graph model etc)
9. Good undestanding on loss fucntions (triplet loss, focal loss, cross entropy, angular, additive angular margin etc) and able to modify to boost in accuracy.
10. Proficiency with image processing techniques, OpenCV, Dlib, OpenMP, parameters tuning, loss convergence, batch processing etc.
11. Ability to design models and losses for unified neural network architecture.
12. Must have knowledge on visualisation tools while training models (Grad-CAM, saliency maps, integrated gradients etc)
13. Proficiency in testing trained-models using confusion matrix, f1-score, bias-variance tradeoffs, roc, auc etc.
14. Must have strong knowledge of deployment strategies like Docker, CI/CD, Test-cases and proficient understanding of code versioning tools (such as GitHub, GitLab, CodeCommit).

Job Location: Work From Any Location
Role: Data Science Engineer
Experience: 2 to 7yrs

Apply for this position

Allowed Type(s): .pdf, .doc, .docx