Equilibrium was founded with a vision for building a company where innovation, collaboration, machine learning, and data science power all aspects of our algorithmic decision-making. We are looking for staff / sr staff machine learning scientists to accelerate the design and delivery of our machine learning models, probabilistic forecasts, and insights dashboards, while helping to shape the science-driven products & processes that will drive the future success of our company.
As a key member of our sciences group, you will play an active role in a) cultivating our culture of experimentation, insights discovery, and incremental delivery, b) facilitating research into state of the art machine learning techniques, c) helping to identify, recruit, train, and mentor members of our growing team of exceptional scientists, and d) partnering with our engineers, product managers, analysts, and commercial team to influence the near to medium term product roadmap.
What you will do
Use research insights to shape product direction: Influence product and engineering roadmaps through presentation of research insights, experimental results, and model performance metrics, in order to evolve organizational direction. Initiate and lead cross-functional engagements to surface, prioritize, formulate, and structure complex and ambiguous challenges where advanced novel deep learning research can have outsized company impact.
Formulate and apply novel machine learning solutions to the energy domain: Tackle complex deep learning & machine learning problems by researching published academic literature, surveying industry techniques & intuition, and executing hands-on experimental testing & modeling. Drive the design, specification, development, and production deployment of our suite of novel deep learning & machine learning solutions. Lead short to medium term research projects that advance the state-of-the-art in deep learning as applied to energy asset management and financial trading.
Performance evaluation: Define and evaluate a suite of success metrics across our portfolio of candidate and deployed machine learning models in order to understand operational characteristics, diagnose sources of under-performance, and identify opportunities for further research & improvement.
The minimum qualifications you’ll need
Passion for clean energy and fighting climate change
An advanced degree in computer science, data science, machine learning, artificial intelligence, operations research, engineering, or related quantitative discipline
4+ years experience in data science, research science, machine learning, or similar role, applying and adapting deep learning, graph neural networks, or reinforcement learning techniques to time series regression & forecasting problems
2+ years experience in the electricity & energy domain (e.g. electricity price forecasting, congestion prediction etc)
3+ years experience with python and the supporting computational science tool suite (e.g. numpy, scipy, pandas, scikit-learn, tensorflow, etc.)
Experience developing, releasing, and tracking performance of ML models in production
Experience communicating mathematical concepts, analytical results, and data-driven insights to both technical and non-technical audiences
A collaboration-first mentality, with a willingness to teach as well as learn from others
Nice to have additional skills
Experience designing and building novel statistical models on time series data, including characterizing probabilistic outcome uncertainty
Experience with dimensionality reduction, component decomposition, or embedding space analysis & visualization techniques (e.g. UMAP, T-SNE, Autoencoder)
Experience with model explainability methods (e.g. SHAP)
Experience with database technologies and sql
Experience with probability, hypothesis testing, and uncertainty quantification
Experience with optimization techniques (e.g. stochastic optimization, robust optimization)
Experience with data visualization and dashboarding technologies (e.g. plot.ly Dash, Streamlit)
Experience leading and mentoring a team of scientists
Demonstrated track record of academic paper or social media publication
