The job requires enhancing and developing the Natural Language Processing capabilities of the core platform. You will be working with other strong NLP data scientists on algorithm design and development to continuously improve and innovate on the state-of-the-art virtual assistants. The work requires significant new thinking and an innovative approach to solve complex problems. You will also be involved in the live product monitoring to understand the NLP model performance.
● NLP algorithms development & enhancement.
● Analyze, prepare algorithm and architecture proposals and solutions.
● Rapid development of demonstrable proof of concepts.
● Unit test case development & unit testing
● Design & code documentation.
● Where required for model development, wrangling large amounts of textual/audio data.
● Supporting development for code that will go to production.
● Retrain and finetune models with new data.
● Monitor models in production.
● Fix issues related to ASR.
● Optimize ASR performance by tuning specific parameters.
● Drive and propose experiments.
● Report results.
● Strong knowledge of common architectures in NLP: language models, embeddings, seq2seq problems, etc.
● Good understanding of metrics, common datasets, and benchmarks.
● 2+ years working on NLP problems.
● Nice to have: Good understanding of Transformers.
● Nice to have: Research experience (both academic or in the industry) modeling NLP problems.
● High proficiency in Python. Experience in Golang is a plus.
● Good understanding of data structures and algorithms.
● Good design and problem-solving skills.
● Experience working with Docker.
● Experience working on cloud environments (especially GCP) is also desirable.
● Familiarity with common development practices (testing, releasing, Scrum, etc.)