Job description

ABOUT BASE Base is America’s next-generation power company. We’re rebuilding the foundation of modern civilization–electricity–by deploying a vast network of distributed batteries that is transforming today’s fragile, centralized grid into a resilient and abundant system. We are engineers, operators, and creatives solving some of the most complex, interdisciplinary challenges of our time. ABOUT THE ROLE As a Software Engineer on the Markets team, you’ll build and automate the decisions our battery fleet and financial commitments run on, from real-time battery dispatch to long-term market participation. You’ll take ambiguous, high-stakes market and control problems and turn them into reliable, automated infrastructure: the simulators, models, and trading-strategy systems behind how our batteries buy, sell, and dispatch power. This role has direct, measurable impact on Base’s top and bottom line. You’ll ship production decision infrastructure with backtest automation and regression-safe releases, deliver measurable improvements to our battery dispatch and control policies, and help bring ML-based methods like model-predictive control and imitation learning into production as we expand into new market modalities and geographies. WHAT YOU’LL DO - Design, build, and maintain the infrastructure, platforms, and algorithmic components we rely on for decisions about battery control and financial market participation — simulators, statistical models, algorithmic trading strategies. - Work closely with utilities, fleet, and market teams to enable our participation in power markets under new modalities and across additional geographies, domestic and international. - Automate the processes the Markets team relies on daily: backtest automation, release processes that prevent regressions in energy arbitrage policy behavior, and iteration speed improvements. - Build the monitoring, alerting, and reporting that tells us how our automated decision processes are performing. - Improve our existing battery physical operation policies. - De-risk, implement, and maintain ML-based methods (supervised learning, model-predictive control, imitation learning for dynamic systems like DAgger) for automated decision making. WHAT YOU'LL BRING - Strong ability to systematize requirements, define success for ambiguous problems with multiple objectives, and propose pragmatic solutions. - A world-class ownership mindset — you own outcomes, not just tasks. - Hands-on experience with AI/ML and HPC infrastructure, from a major cloud provider (AWS, GCP, Azure) or specialized ones (Modal, Beam). - Strong interest in applied math and machine learning methods. - Experience solving MLOps or similarly challenging observability/monitoring/versioning problems. - Nice to have: experience with decision-making under uncertainty for stateful systems. ABOUT THE TEAM Base’s Markets team owns the systems that decide how our fleet of batteries buys, sells, and dispatches power across every market we operate in. Every dollar of revenue and every megawatt of grid support we deliver runs through a decision this team automates — from real-time battery dispatch to long-term financial commitments. As we expand into new markets and geographies, the fundamentals get harder: more uncertainty, more constraints, more money on the line. This team gets to define, from scratch, how a distributed battery fleet should behave in power markets that were never built for it, with direct, measurable impact on Base’s top and bottom line. Please note: Base is a startup, which means priorities shift and evolve quickly. Your role may expand or change based on the needs of the business at any given time, so the responsibilities listed may not be exhaustive. OUR VALUES - First Principles Thinking: Question assumptions. Principles > rules. - Operate at Base Pace: Focus on what matters, act quickly, and learn by doing. - Give & Get Feedback: Be direct, be humble, and maintain a growth mindset. - Everyone’s an Owner: Follow through on commitments and own results. - Strong Opinions, Loosely Held: Drive clarity and make calls with imperfect information. - Committed to the Mission: Rebuilding the grid is a big challenge. We work hard because we care deeply about the impact we’re creating. We work in-person. It’s not a 9-to-5. We are all-in. - Fun & Optimism Coexist with Grit: Collaboration and celebration coincide with the intensity of building real things. DO THE BEST WORK OF YOUR LIFE AT BASE.

Requirements

Must-haves

Work closely with utilities, fleet, and market teams to enable our participation in power markets under new modalities and across additional geographies, domestic and international.
Automate the processes the Markets team relies on daily: backtest automation, release processes that prevent regressions in energy arbitrage policy behavior, and iteration speed improvements.
Build the monitoring, alerting, and reporting that tells us how our automated decision processes are performing.
Improve our existing battery physical operation policies.
De-risk, implement, and maintain ML-based methods (supervised learning, model-predictive control, imitation learning for dynamic systems like DAgger) for automated decision making.
Strong ability to systematize requirements, define success for ambiguous problems with multiple objectives, and propose pragmatic solutions.
A world-class ownership mindset — you own outcomes, not just tasks.
Hands-on experience with AI/ML and HPC infrastructure, from a major cloud provider (AWS, GCP, Azure) or specialized ones (Modal, Beam).
Strong interest in applied math and machine learning methods.
Experience solving MLOps or similarly challenging observability/monitoring/versioning problems.
Please note: Base is a startup, which means priorities shift and evolve quickly. Your role may expand or change based on the needs of the business at any given time, so the responsibilities listed may not be exhaustive.

Nice-to-haves

Nice to have: experience with decision-making under uncertainty for stateful systems.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

mlaiawsgcpazuremachine learning

Platform: ashby
Effort: quick