In this project, Luxoft helps the customer (German OEM) develop, train, and deploy Machine Learning models for internal customer needs. The models include:
- Automatic Warranty Claim Approval
- Financial Forecasting
- Automatic anomaly detection in engine signals (IoT on mining trucks)
All models use data in the tabular form as an input (i.e. no computer vision).
The customer develops engines for mining trucks, yachts, cruise ships, power generators, etc.
Five data scientists and a product owner.
Python, unit testing, pandas, numpy, Machine Learning, Data Science, Azure Databricks, MLFLow, Docker, Gradient Boosting Trees (LightGBM), Pyspark, Tensorflow, Keras, Deep Learning, LSTM, Kanban, Azure DevOps.
Working with a German OEM on projects that really go into production and bring value to the customer. Working with a range of Data Science and Machine Learning technologies as well as diverse problems.Responsibilities:
- Build, train, run, own and validate predictive and explorative analytical AI models
- Analyze large and complex data sets to derive valuable insights
- Write and develop production level code/pipelines
- Collaborate with DevOps engineer and Data engineer to bring models to production
- Actively monitor AI models during their lifecycle and intervene in case of trouble
- Presentation of qualitative and quantitative results obtained during development and running of these models
- Delivery of fast, stable, scalable, robust and future-proof AI microservices