ShoptakiSoftware Engineer
Jun. 2024Engineered robust facial recognition algorithms using Python, TensorFlow, and OpenCV, achieving a 13% improvement in F1-score accuracy through rigorous unit and integration testing across the full SDLC.
Designed and deployed real-time liveliness detection and facial recognition APIs with FastAPI, ensuring high-availability and resilience in distributed systems.
Developed comprehensive audio processing pipelines with pydub, torchaudio, and custom effects for normalization, noise reduction, and format conversion, enhancing ML model input quality.
Leveraged the microsoft/wavlm-base-plus-sv model with PyTorch and Transformers to generate normalized voice embeddings for