Senior Applied AI Engineer, Retrieval and Semantic Systems

At Instructure, we believe in the power of people to grow and succeed throughout their lives. Our goal is to amplify that power by creating intuitive products that simplify learning and personal development, facilitate meaningful relationships, and inspire people to go further in their education and careers.

We do this by giving smart, creative, passionate people opportunities to create awesome.

And that's where you come in:

Our team builds AI-native capabilities, reusable AI systems, and shared infrastructure that power multiple products and workflows across the platform.

We are looking for a Senior Applied AI Engineer to own retrieval and semantic systems end to end. This role builds and operates production retrieval as a core capability, including the retrieval infrastructure layer (indexing, storage, scaling, cost, and reliability), quality evaluation, and iteration loops that improve relevance over time. You will partner with platform engineers on deployment standards and observability, but you will own retrieval architecture decisions and day-to-day operation.

You will work closely with product, engineering, and research partners to turn advanced AI ideas into reliable product capabilities used at scale.

What You’ll Do

• Design, build, and ship production retrieval systems that power AI product capabilities

• Own vector store selection and operation, including scalability, latency, reliability, cost, and multi-tenant design

• Build indexing and refresh pipelines (chunking, embedding generation, backfills, deletes, versioned indices)

• Implement semantic retrieval patterns, including embeddings, similarity search, metadata filtering, and reranking

• Define and run retrieval evaluation: gold sets, offline metrics,, slice analysis, drift detection, and regression gates

• Partner with platform engineers on CI/CD, service templates, monitoring, and incident readiness while owning retrieval correctness and evolution

What You’ll Need

• 6+ years of experience building and shipping production ML or applied AI systems

• Proven experience owning a retrieval system in production, including vector store selection and operation

• Strong Python engineering skills and experience building services/APIs (for example, FastAPI or similar)

• Hands-on experience with embeddings, approximate nearest neighbor search concepts, and retrieval or ranking systems

• Experience designing indexing and refresh strategies, including data quality controls and safe backfills

• Strong judgment on tradeoffs across relevance, latency, cost, and operational complexity

• Strong communication skills and ability to collaborate across engineering, product, and research teams

It Would Be a Bonus If You Had

• Experience with hybrid retrieval (lexical plus vector), learning-to-rank, or domain-specific reranking

• Experience with graph-structured context systems or knowledge graph integration

• Experience building evaluation and observability for LLM or retrieval systems (quality drift, failure analysis, regression prevention)

• Experience with AWS-native architectures for retrieval and indexing services

• Experience in education technology, content, curriculum, or skills modeling

Growth & Impact - In This Role, You’ll Be Expected To

In this role, you will define retrieval as a core capability across multiple AI products. You’ll establish the standards and evaluation loops that make semantic systems reliable, measurable, and scalable in production.

Why Join Us

Join us and help shape the future of education by turning cutting-edge AI into reliable product capabilities.

At Instructure, we’re on a mission to help educators and students learn together, anytime, anywhere, and however works best. You’ll join our research-driven team tackling education’s biggest challenges with cutting-edge technology. Our projects have included making sense of unstructured feedback, applying large language models to save teachers’ time and improve student experiences, classifying partner networks for smarter recommendations, and detecting fraud to protect resources for real learners.

We value diversity, creativity, and passion, and invest in our teams through mentorship, hack weeks, internal conferences, and a culture where innovation thrives. Here, you’ll have the chance to build the next generation of LMS features that make a real impact on students and teachers, and do it in a collaborative, supportive environment that encourages experimentation and growth.

Get in on all the awesome at Instructure!

We offer competitive, meaningful benefits in every country where we operate. While they vary by location, here's a general idea of what you can expect:

• Competitive compensation, plus all full-time employees participate in our ownership program - because everyone should have a stake in our success.

• Flexible work culture. Our remote, hybrid and in-office collaboration spaces vary by role, team and location.

• Generous time off, including local holidays and our annual “Dim the Lights” period in late December, when teams are encouraged to step back and recharge based on departmental needs.

• Comprehensive wellness programs and mental health support

• Annual learning and development stipends to support your growth

• The technology and tools you need to do your best work — typically a Mac, with PC options available in some locations

• Motivosity employee recognition program

• A culture rooted in inclusivity, support, and meaningful connection

We believe in hiring great people and treating them right. The more diverse we are, the better our ideas and outcomes.

Instructure is an Equal Opportunity Employer. We comply with applicable employment and anti-discrimination laws in every country where we operate.

All employees must pass a background check as part of the hiring process. To help protect our teams and systems, we’ve implemented identity verification measures. Candidates may be asked to verify their legal name, current physical location, and provide a valid contact number and residential address, in accordance with local data privacy laws.

Any attempt to misrepresent personal or professional information will result in disqualification.