A leading energy provider in the APAC region is looking to moving from traditional energy/utility industry to a more data-driven company and create data-driven products.
The client is looking for someone who can discover patterns hidden in large amounts of data and make decisions from different sources. The primary focus is on applying data wrangling and machine learning techniques to build high quality anomaly detection, prediction and recommendations systems integrated in their products.
- Understand business logics from domain experts and come up with reasonable targets for data projects
- Data fetching from different sources such as database, big data lake running on hadoop/hive
- Enhancing data by building autonomous pipelines from different sources
- Data wrangling by preprocessing, cleansing, and feature engineering
- Applying state-of-art machine learning techniques such as RNN, CNN for predictions and anomaly detections
- Build agile data products in a team of data engineers, scientists and business users
- Doing ad-hoc analysis and presenting results in clear manner
- Guide junior team members on their projects
- Help find opportunities from different business partners
- Experienced with common data science toolkits such as Python/R.
- Understand machine learning models, pros and cons.
- Strong experience in data visualization tools such as D3.js, matplotlib, and etc.
- Good understanding of statistics, such as distributions, A/B testing, model over/under-fitting.
- Experience with one of the deep learning libraries such as Tensorflow, Keras, Pytorch, CNTK, MXNet, and etc.
- Masters’ or PhD of Computer Science/Engineering, Applied Maths or other engineering related area.
Please contact Mabel Ngu at +65 6950 0366 or firstname.lastname@example.org for a confidential discussion
EA License no: 16S8066 | Registration no: R1980996
Only successful candidates will be notified.The incumbent is required to be able to work with multiple levels of the organization to understand the business requirements, build solutions for different end-users and has strong machine learning (time series) capabilities.