Project Overview
Myst AI is a startup that uses machine learning to help energy companies better predict the future and achieve their goals for a carbon-free energy grid. One of the last advantages that dirty energy has over distributed clean energy is that dirty energy is more predictable; we aim to close that gap.
We work with leading clean energy companies in the US and Europe to create more accurate, reliable, and scalable forecasts. Our clients include community choice aggregators and nonprofits that work to build low-income, carbon-free housing addressing the intersectionality of the climate crisis.
Based on our strong software and sales, Myst recently raised $6 million from a social impact venture capital firm.
Technical Work
As the lead software engineer for the company, I helped design our core software architecture and data model, including work we patented on our design for networked time series forecasting. I helped design and develop the Python back-end application, helping refactor the codebase and implement pioneering platform features. I personally built our entire Javascript front-end application and I also migrated our task processing infrastructure to run entirely on Kubernetes. I led our observability implementation efforts across a wide range of metrics and developed most of our best practices and processes for code quality, on-call schedules, incident response, and developer tooling.
As the acting engineering manager for the team, I supported my direct reports with their growth plans and daily work. I also developed our internal processes for organizing sprints, objectives and key results (OKRs), post-mortems, and retrospectives. I led our engineering hiring efforts and also helped with hiring for the data science team.
I also helped establish the Diversity, Equity, and Inclusion (DEI) committee at the company, and 100% of non-executive staff quickly joined the committee. Through authentic and vulnerable conversations and collaboration, we embedded DEI and social impact goals part of the company’s core OKR framework so that it's become a core part of Myst's daily focus.
Our Javascript front-end application was built using React, Redux, Typescript, Material Design Components, Styled Components, and more. Our Python back-end application is built using Flask, Pandas, Tensorflow, XGBoost, Scikit-learn, and more.
Our infrastructure is hosted in Google Cloud Platform and includes a wide variety of services. Our codebase is thoroughly unit tested (in the front and backend) and our CI and CD pipelines are fully automated.