Lynker Technologies, LLC has an exciting opportunity for a full-time Software Engineer - Machine Learning & Artificial Intelligence located in Tuscaloosa, AL
The NOAA National Water Model (NWM) provides operational analyses and assessments of the full range of water cycle components across the United States, with an emphasis on streamflow. Accurate simulation of these elements entails proper representation of all relevant physical processes and anthropogenic influences that impact the flow of water. While natural factors work in a complex and interwoven fashion to affect streamflow, reservoirs, which are operated in ways that are governed by a mixture of natural forcing and water law, add yet another layer of complexity. Reservoirs can be operated for flood control, water supply, energy production, and water storage purposes, with each purpose having unique governing requirements. Over 1,000 reservoirs are represented in the current version of the NWM. However, the current representation is rudimentary, with each reservoir modeled in a passive level-pool fashion. Enhancements to representation of reservoirs in the NWM will need to take into account the wide variety of factors that influence operations over the course of each water year. For reservoirs that feature an existing set of operating rules, this may simply involve integrating those rules into the NWM via a new module. However, for the majority of reservoirs where operating rules do not exist, a different approach will be needed. This position will work on ML/AI approaches for doing this. NOAA believes such techniques have the potential to intelligently integrate a wide variety of factors, from water law to antecedent/forecast conditions, to past behavior, and form a new set of rules by which to model reservoir operations within the NWM.
This position will work with a hydrology expert to design, construct, integrate and test new NWM modules necessary to achieve this enhanced capability--in particular it will replicate the behavior of reservoir managers in order to accurately simulate reservoir releases within the NWM framework. The position will develop documentation and code to accomplish the reservoir modeling improvements outlined above. They will be responsible for integrating the code into the current version of the NWM.
- Strong background in Machine Learning (ML)
- Solid background in Large Scale Data Mining and Artificial Intelligence (AI)
- Strong Experience using appropriate open source frameworks to understand/enhance the inner workings of an AI framework of choice (e.g. TensorFlow, MXNet, Theano, Caffe etc.)
- Delivery of a previous AI based project with success and failure stories from a real customers, with a good sense of lessons learned and a clear interest and passion in trying out what was learned in new projects
- Strong hands-on coding expertise with C/C++, Python, Scala, Lua, MATLAB, or any proficient AI language of choice
- A mathematical, statistical and probability inclination and deep understanding of some AI concepts
- A sense of ambition and passion to work in the environmental sciences (water resources management) using ML/AI
- MS with 5+ years in Computer Science field, or PhD with 2+ years in Computer Science field, or BS with 8+ years of applied machine learning experience.
- Degree preferred, not required if candidate possesses equivalent applied machine learning experience and expertise
- End to end hands-on ownership of machine learning systems deployable across various projects including data pipelines, model generation, and training and inference engines
- Algorithm development, Model enhancements, around key NWM research areas (innovation areas) that NOAA is focusing on. This will require paper reading, and staying ahead of the game by knowing what is and will be state of the art in the ML field
- Must have a collaborative nature and willing to work with Earth Scientists and other Software Engineers (comprised of a mix of employees and consultants) onsite.
Closing Date: 30th September 2017
Applications by email to firstname.lastname@example.org