Careers · Software Engineering
Applied AI Engineer
Builds the shared capabilities that migrations, studies and customer operations draw on.
About the role
What this role is for.
GridStrong's Dashboard reads models the way a consultant would, finds the right relay file among a plant's thousands of documents, extracts settings from OEM tables, drafts assessment notes and tracks every operator notice as a task. Those capabilities are built once and used everywhere: by the migration tooling, the study engine and the customer operations team. The Applied AI team owns them and decides which work runs in production code and which runs through operator-in-the-loop import and export processes.
You will build those shared capabilities. The work is engineering, not research: reliable, repeatable extraction and classification over messy engineering documents, with validation that proves the output, and the plumbing that makes the same capability available to every team that needs it.
Team: Software Engineering
Level: Engineer
Location: Remote, United States or Canada
Works with: The VP, Applied AI and the Applied AI team, Forward Deployed Engineering, Power Systems Engineering, and the Forward Deployed Managers and Power Systems engineers whose daily work runs on these capabilities.
What you will do
- Build and maintain shared extraction and classification services for documents, relay settings files, OEM tables and operator correspondence.
- Build the automated assessment capabilities that draft study notes and pinpoint values from simulation results, with accuracy measured against engineer review.
- Build evaluation harnesses: ground-truth sets, regression tests and accuracy reporting for every capability in production.
- Unify how customer operations tooling and production systems call these capabilities, and decide with the VP which work belongs in each.
- Pair with Forward Deployed Engineering and Power Systems Engineering to productize what they have built by hand.
- Monitor cost, latency and failure modes in production and fix them.
- Document capabilities so engineers and operations staff use them correctly.
What we are looking for
- Four or more years of software engineering in Python, with production systems you have run.
- Experience building extraction, classification or language model pipelines over real documents, with the evaluation discipline to prove they work.
- Strong data engineering habits: schemas, validation, provenance and reproducibility.
- Comfortable in a Ruby on Rails codebase alongside Python services.
- Able to work with engineers who are the domain experts and translate their judgment into testable logic.
- Clear written communicator.
Nice to have
- Energy, utility or other regulated domain experience.
- Experience with PDF and engineering drawing extraction.
- Experience operating model-backed services at scale: caching, batching and cost controls.
The first 90 days
- You have shipped an improvement to a shared capability and measured its accuracy change against ground truth.
- You own the evaluation harness for one capability in production.
- You have paired with Forward Deployed Engineering or Power Systems Engineering to productize one hand-built step.
Apply for this role
Send a note and whatever shows your work best to careers@gridstrong.ai. Tell us what you would want to build first. We read every note and reply within a week.