Principal Engineer, AI

Role Description

You are a Principal Engineer who is both AI-native and relentlessly product-minded. You turn ambiguous, cross-team opportunities into clear technical direction, and then you prove the path by building. You care about the flagship experiences you ship, and you obsess over the shared "paved highway" that makes it easy for the teams around you to ship safe, reliable AI over and over again.

You will succeed in this role if you are:

• Driven by outcomes. You care about helping centers stay full, save hours every week, and serve families better, not shipping "an AI feature" for its own sake.
• The strategic architect and prolific builder. You bridge strategy and deep implementation. You set the technical direction for your organization, then you get hands-on to deliver reference implementations that teams can adopt with confidence.
• Decisive on build vs. buy. You have sharp judgment on when to leverage off-the-shelf models and platforms versus when to invest in proprietary infrastructure, fine-tuning, data flywheels, or evaluation systems to create durable advantage.
• Velocity as a strategy ("show, don't tell"). You move with the hunger of a founding engineer: you prototype boldly, de-risk quickly, and use working systems, not slides, to align teams and raise their ambition and pace.
• A force multiplier with a platform mindset. You build the paved road: evaluation harnesses, retrieval and context patterns, safety guardrails, observability, and developer experience that let product and engineering teams ship AI safely and autonomously.
• Deep on fundamentals, modern in practice. You understand why models fail (data, evaluation, human feedback, reliability), and you aggressively use modern AI-native tooling to iterate faster without sacrificing rigor.
• Security- and trust-minded. You treat sensitive school, educator, and family data with care, design for least-privilege access, and build systems that are explainable, monitorable, and resilient.
• Mission-aligned. The idea of using AI to expand childcare capacity and improve early education outcomes, at meaningful scale, genuinely motivates you.

What You’ll Do

You will make brightwheel measurably more valuable to schools, centers, educators, and families by delivering AI-powered capabilities that move business outcomes—for example retention, enrollment conversion, payment success, operational efficiency, and support burden. This is not a research role or a "build cool demos" role; success is defined by impact in production.

• Own outcomes end-to-end. Take responsibility for turning high-leverage opportunities into shipped, adopted capabilities with clear measurement, iteration loops, and sustained improvement—not one-time launches.
• Define the AI technical direction for your organization. Translate customer problems into a practical strategy, reference architectures, and a roadmap that multiple teams can execute against, with minimal guidance.
• Drive build-versus-buy decisions. Recommend when to partner on models and tooling versus when to build differentiated capabilities (data flywheels, evaluation systems, workflow-specific intelligence), backed by evidence from fast, hands-on spikes.
• Lead by implementation. Rapidly prototype and ship reference implementations to prove feasibility, de-risk decisions, and raise your teams' expectation for speed and ambition.
• Help build the paved highway. Contribute shared foundations—evaluation harnesses, retrieval and context patterns, safety guardrails, observability, and developer experience—so teams can ship reliable AI features repeatedly and independently.
• Raise the bar across your organization. Mentor senior and staff engineers, lead design reviews for your teams, help establish standards for quality and safety, and align stakeholders to drive adoption.

Qualifications

• Foundational AI Depth: You possess deep intuition for why models fail, gained from experience that predates the LLM boom. You have likely worked with model training, fine-tuning, or classical NLP/ML, giving you the grounding to make sound architectural decisions.
• 8+ Years of Engineering Excellence: You have end-to-end ownership of large, business-critical systems. You launch large-scale and/or highly complex software that requires members of multiple teams to contribute, and you deliver largely independently with minimal guidance.
• Architectural Strategy: Experience formulating architectural strategy at the organization level. You have designed systems that required execution across multiple teams, and you regularly influence technical and product priorities.
• Mentorship and Hiring: A track record of mentoring senior and staff engineers and improving their skills, knowledge, and ability to get things done. You actively interview engineers and your judgment carries weight in debriefs.
• Production AI at Scale: A proven track record of shipping AI-powered products to production. You understand the "last mile" of AI—evaluation, monitoring, and safety—and have built the tools that allow teams to sleep soundly at night.

Nice-to-haves

• Experience earlier in your career with model training, fine-tuning, classical machine learning, or natural language processing—enough to have intuition for why models fail and how data and evaluation shape outcomes.
• A portfolio of real work (open-source, demos, writing, talks, or shipped side projects) that shows taste, velocity, and how you think about applied AI systems end-to-end.
• Experience building shared internal platforms or frameworks (for example evaluation services, retrieval infrastructure, policy and safety guardrails, observability tooling) that became the default path for multiple teams.
• Formal training in computer science (4-year CS degree or equivalent depth in core CS topics).
• A strong bar for operational excellence: secure-by-default design, performance and cost discipline, testability, incident readiness, and a track record of improving development, testing, and on-call practices for complex systems.

Technology

• Backend: Ruby on Rails, Sidekiq
• Data: PostgreSQL on Amazon RDS, Redis, and event and analytics pipelines
• Frontend: React with TypeScript and Emotion
• Mobile: Native iOS (Swift) and Android (Kotlin with Jetpack Compose)
• Cloud & Infrastructure: Docker, Kubernetes on Amazon EKS, GitHub Actions and FluxCD for CI/CD, and AWS services such as S3, CloudFront, CloudWatch, and SNS
• AI & Automation: AWS Bedrock and other hosted large language models, orchestration and agent frameworks, and modern AI coding tools like Cursor, Claude Code and Codex