Qualification :
Work Location : Bangalore
Work mode : On site
Notice period : Immediate to 30 days
Experience : 2 - 6 years
Responsibilities :
- Build Audio ML / DL models for given requirement
- Build models from scratch wherever required
- Must be able to code relevant research papers independently
- Tune to improve accuracy and latency
- Support to satisfy the deployment requirements of the product
- Apply transfer learning wherever required
- Training, fine tuning and optimization of different flavors of transformer models
- Hands on experience with LLM, fine tuning and Optimization of LLM
SKills :
Audio AI concepts, ML / DL working experience in Pytorch / Tensorflow / Keras frameworkVariations of CNN, RNN, Attention Mechanism, Transformers and large models like wav2vec, whisperProficiency in machine learning / audio pre-processing libraries / frameworks such as Librosa, Kaldi, sklearn, python speech features, Numpy, Scipy, Pandas, Matplotlib.Signal Processing fundamentals, understanding the data characteristics and wave signalsTransfer Learning Experience in Acoustic models for speech recognition systems.Experience in Voice activity detection with noise analysis & denoisingExperience in WakeupWord, ASRWorking experience of LLM fine tuning, optimization and performance improvement.