Staff Machine Learning Engineer
Service Intelligence Platform is responsible for a wide variety of NLU, Search, Recommendation, and speech related products supporting our stakeholders in Data Science and Service Engineering teams. Our products are geared towards helping customers and agents be a part of world-class post order experience, and include goal-oriented Virtual Assistants, ASR, Q&A systems, and Recommendation in Search.
The projects that our teams work on are built from the ground up – we look for entrepreneurial individuals who want to take ownership over their own agenda and thrive in a collaborative team environment.
What You'll Do:
- Deploy and maintain production models using both internal and external solutions
- Improve the pace of innovation and experimentation by introducing best practices and tools for Data Science workflow and DevOps
- Help grow other engineers by mentoring and developing learning opportunities
- Deal with rapidly evolving business priorities.
- Collaborate with some of the best engineers and data scientists in the industry.
What You Will Need:
- A strong statistical background in ML, NLP, deep learning models along with familiarity in probabilistic information retrieval and optimization methods
- Professional experience of building and deploying ML/AI apps to production, in realtime and Batch mode at scale
- Strong understanding in data structures, algorithms and software design concepts in Streaming, REST, gRPC
- Expert knowledge in one of Java/Scala/C++
- Minimum of 2 years experience with scripting language Python, & Docker
- Hands-on distributed systems compute experience using Spark/Dataproc, Kafka consumers, Hive or equivalent
- Good understanding of Tensorflow, Pytorch, Pandas, Scikit-learn and Airflow
- Strong understanding in model inferencing lifecycle, Monitoring, feedback loop and data capture in real time at scale
Nice to Have:
- Experience with one of GCP/AWS and Kubernetes or Swarm
- Experience with Dataproc, MLflow, Horovod, Michelangelo
- Experience with semi-supervised learning, BERT, XGBoost
- multi-arm bandits setup experience
- UniMRCP, RTP understanding at a high level
- WebSocket understanding on a high level
- Search Engine exposure