Staff AI Research Engineer

About the role

You will own the design and execution of AI research initiatives that directly enhance learning outcomes and business performance within the Duolingo ecosystem. This role requires you to architect and implement advanced machine learning systems that operate at scale for a global user base. You will own the full lifecycle of intelligent products, from ideation and data strategy through deployment and continuous optimization in production environments. The position demands deep collaboration with product and engineering stakeholders to translate ambiguous product challenges into robust, data-driven AI solutions. You will establish and refine evaluation frameworks to measure the real-world impact of AI models on user engagement and monetization. Ownership of cutting-edge techniques in large language models and sequential decision-making is essential to guide the team toward best-in-class innovation. Ultimately, you will shape the technical vision for the Monetization group, ensuring that research breakthroughs are reliably translated into high-impact, revenue-positive educational experiences.

Key facts

What you'll do

Lead the discovery and definition of complex product opportunities requiring advanced AI techniques, translating ambiguous requirements into concrete research agendas and technical specifications.

Architect, implement, and iterate upon large-scale machine learning pipelines, with a primary focus on developing and optimizing bandit models and large language model applications for monetization.

Design and execute rigorous offline and online evaluation strategies, including A/B tests and multi-armed bandit experiments, to quantitatively validate model performance and business impact before full rollout.

Partner closely with frontend, backend, and data engineering teams to integrate research prototypes into production-grade services that handle Duolingo's massive user traffic reliably and efficiently.

Perform deep feature engineering and data curation to construct high-quality training datasets that drive optimal model behavior across diverse learner segments and content types.

Champion experimentation culture by running and analyzing hundreds of concurrent experiments, interpreting results, and providing actionable recommendations to influence product strategy at the leadership level.

Conduct thorough quality assurance and monitoring of deployed models, ensuring they meet strict standards for reliability, fairness, and performance in live educational environments.

Mentor and uplevel junior engineers and researchers, fostering a collaborative environment that encourages knowledge sharing, technical excellence, and career growth within the AI research community.

Stay at the forefront of academic and industry research in machine learning, particularly in reinforcement learning with large language models, and proactively propose novel applications to maintain Duolingo's competitive edge.

Own the end-to-end delivery of AI projects, managing timelines, dependencies, and stakeholder communication to ensure successful deployment of solutions that meet business objectives and user needs.

Requirements

You must possess a Bachelor's, Master's, or PhD degree in Computer Science, Engineering, or a closely related field with a focus on machine learning or artificial intelligence.

You must have proven professional experience as an AI Research Engineer or similar role, with a strong track record working on large language models, multi-armed bandits, and related sequential decision-making systems.

You must have extensive hands-on experience across the full machine learning stack, including feature engineering, training data development, model fine-tuning, reinforcement learning, quality evaluation, deployment, and production monitoring.

You must demonstrate advanced proficiency in modern machine learning frameworks and libraries, enabling you to build scalable and efficient models from concept to production.

You must have excellent written and verbal communication skills, with a proven ability to influence and lead technical discussions among cross-functional stakeholders.

You must have strong leadership capabilities and the experience to manage, mentor, and inspire a team of engineers and researchers toward shared technical goals.

You must maintain a deep understanding of machine learning theory, algorithms, and best practices, applying this knowledge to solve complex real-world problems effectively.

You must have a history of delivering high-quality, scalable software solutions that are robust, maintainable, and performant in high-traffic production settings.

Nice to have

Experience publishing research at top-tier academic conferences or journals in machine learning or related fields is preferred.

Familiarity with educational technology platforms and learning science principles is a preferred qualification for this role.

Practical notes

This role is based in Pittsburgh, Pennsylvania, requiring relocation or local candidacy for in-office work.

Employment is full-time, and the position is exempt from standard work hour constraints as needed to meet business demands.

Visas may be sponsored for qualified candidates, subject to eligibility and role requirements.

The salary range displayed reflects the minimum and maximum target base salary for this position in the United States.