EdTech engineering capabilities
The standards your customers run on, and how AI has to fit them.
AI features in education only work when they fit the LMS, SIS, rostering and sign-in systems schools already use. These are the standards we scope against in a Product Review and offer as integration work in your build.
Standards and systems
Integration services we offer.
Each is an integration service we scope in the Product Review against your product's existing integrations and the published specification, and deliver in your codebase.
| Area | Standards | What it enables |
|---|---|---|
| LMS integration | LTI 1.3 / LTI Advantage (deep linking, assignment and grade services, names and roles) | Launch an AI feature from inside the institution's LMS, pick content into a course, and write scores back to the gradebook. |
| Rostering | OneRoster 1.2, Clever, ClassLink | Classes, enrollments and roles arrive from the district's source of truth, so the AI only sees the classes a teacher actually teaches. |
| Student information systems | Ed-Fi, SIS vendor APIs | Read the records a feature needs, at the permission level the user already has, without copying the SIS. |
| Learning records | xAPI, cmi5, SCORM 1.2 / 2004 | Record practice sessions, scores and completions in a learning record store or the LMS, in a format customers already report on. |
| Sign-in and provisioning | SAML 2.0, OpenID Connect, SCIM | Single sign-on with the institution's identity provider and automatic account provisioning and removal. |
| Content and assessment | QTI, Common Cartridge, CASE | Import and export questions, course packages and standards alignments so AI-drafted content fits existing libraries. |
| Accessibility | WCAG 2.1 AA, Section 508 | Keyboard use, contrast, labels, captions and screen-reader testing on every new screen, reviewed against WCAG 2.1 AA. |
| Privacy design | FERPA, COPPA, state student-privacy laws | Data minimization, institution isolation, age-appropriate design for users under 13, and no training on student data. |
How the team works
One team across product, AI, platform and quality.
A named lead engineer and US-hours overlap, from the first Product Review to production.
Product and learning design
Business analysts who map the workflows of every user your product serves (teachers, learners, admins, parents) before anything is built, and design AI that saves work people already do.
AI engineering
Retrieval with citations, voice and real-time AI, document extraction, text-to-SQL and agents with approval steps, evaluated on test sets your team reviews.
Platform engineering
Multi-tenant SaaS, role-based access, Stripe billing, cloud on AWS, Azure or GCP, and CI/CD with approval gates.
Quality and accessibility
Regression testing, accessibility testing and release management, so AI features ship without breaking what already works.
See it on your own product.
Bring one product, workflow or AI feature. In 30 minutes we tell you which of the three ways to start fits, what it would cost to serve and what your customers' security review will ask.
Calls run in US Central time. No generic AI presentation: we discuss your product, users and production constraints.