We seek a Machine Learning Engineer to support development of a suite of valuation models and data products. This person would work alongside software engineers building the technology to manage our data, and the data scientists conducting statistical analysis and developing models. As a Machine Learning engineer, you would complement these efforts by demonstrating how ML can enhance approaches to data discovery, feature engineering, and predictive modelling.
What you'll be responsible for
- Develop proof-of-concept ML models to solve problems for a variety of data and domains, which might include but not limited to:
- -text extraction and classification from millions of pages of technical reports;
- -estimating building density from satellite images; or,
- -training ML models on semantically related public records.
- Explain the effectiveness and assumptions of your approach and guide collaborative research and methodology development with appropriate technical rigor.
- Build predictive models in production at scale.
- Contribute to decisions about our technology stack, particularly as it relates to end-to-end ML model and data flow and automation.
- Stay current with latest ML algorithms and methods and share knowledge with colleagues in data science and engineering and externally as appropriate.
What we're looking for
- Critical professional skills include: a curiosity to discover new approaches to problem-solving, a drive to initiate ideas and collaborate with colleagues to see them through, and an ability to communicate technical material clearly.
- Skills in Python or R (+Scala a bonus), including ML libraries (e.g. SKLearn, NumPy, SciPy), SQL, parallelization and tools for large scale computing (e.g. Spark, Hadoop), matrix algebra, and vectorization
- Experience with at least one of the DL frameworks (e.g. PyTorch, Caffe, TensorFlow, Theano, Keras) and a perspective on what distinguishes them
- Experience with supervised and unsupervised learning algorithms, including cluster analysis (e.g. k-means, density-based), regression and classification with shallow algorithms (e.g. decision trees, XGBoost, and various ensemble methods) and with DL algorithms (e.g. RNN w/ LTSM, CNN)
- Experience with advanced methods of ML model hyper-parameter tuning (e.g. Bayesian optimization)
- Deep understanding of statistics and Bayesian inference, linear algebra (e.g. decomposition, image registration), vector calculus (e.g. gradients), time series analysis (e.g. Fourier Transform, ARIMA, Dynamic Time-Warping)
- Proven track record of building production models (batch processing at minimum, online ML a bonus)
- Experience in at least one of the following data domains: highly disparate public records, satellite images, text in various states of structure
- Experience with remote computing and data management (e.g. AWS, GCP suite of tools)
Bonus points for
- Being a technical lead on a successful ML-based product
- Doing applied research (at graduate school level or equivalent) with: 1) Geospatial analysis, 2) Image processing (e.g. denoising, image registration), 3) NLP, or 4) ML with semantic databases
- Experience with streaming models or online ML
- Authored technical publications or presented work in the field
- Domain expertise in real estate or the built environment
What we give you
- You will have the opportunity to accelerate our rapidly growing organisation.
- We're a lean team, so your impact will be felt immediately.
- Agile working environment with flexible working hours and location, career advancement, and competitive compensation package
- GeoPhy is a family friendly company. We arrange social activities to help our employees and families become familiar with each other and our culture
- Diverse, unique colleagues from every corner of the world
by via developer jobs - Stack Overflow
No comments:
Post a Comment