LMS Bridge reads assessment results from your LMS, pinpoints each student's misconception from their actual answers, and runs a guided AI tutoring session that helps them recover — before the gap compounds. It installs into Blackboard, Brightspace, Canvas, and Moodle with one standard. It's free and open source (AGPL-3.0) — run it on your own infrastructure, with your own AI key.
No login needed — try the real tool inside your LMS. Pick one:
In cumulative STEM courses, a shaky grasp of binary arithmetic or conditional probability in week 3 turns into failure in week 10. Instructors have rich gradebook data but no way to act on it, student by student, at the moment it matters.
For multiple-choice items, the exact wrong option a student picks reveals a specific misconception. LMS Bridge reads it and targets that.
A pedagogically-constrained AI tutor runs a Socratic session the student completes — it never does the work for them. Not a cheating tool.
Every at-risk student gets individualized remediation automatically, the moment a formative assessment is graded.
A complete adaptive-learning layer that augments — never replaces — your course.
Per-question answer analysis maps each distractor to the misconception it reveals.
Turn-by-turn Socratic dialogue, grounded in the student's own mistakes, that they complete.
Upload your PDFs, notes, and problem sets; the tutor uses your notation and examples.
Class concept-risk, per-student drill-down, rubric breakdowns, and CSV export.
Enable or disable any assessment as an adaptive input, and recompute on demand.
FERPA-aware, role-based access, and a bring-your-own-model option for full data governance.
From inside your existing course — no separate logins, no new gradebook.
Your LMS admin registers LMS Bridge as an LTI 1.3 tool. It then appears inside any course.
LMS Bridge reads scores (and, for quizzes delivered through it, per-question answers) and updates a private mastery estimate per concept.
When a concept drops to at-risk, a tailored AI tutoring session is created for that student.
Students complete the guided session; instructors get class- and student-level insight.
One integration, every platform. LMS Bridge is built on LTI 1.3 / LTI Advantage, the universal standard from 1EdTech — so it installs the same way into Blackboard (Anthology), Brightspace (D2L), Canvas (Instructure), Moodle, and Schoology.
Students and instructors launch from inside the course — no extra password (LTI Core + OIDC).
Names & Role Provisioning Services (NRPS) pulls the class list and roles for you.
Assignment & Grade Services (AGS) reads assessment outcomes; mastery can post back as a non-graded column.
Deep Linking lets instructors drop LMS Bridge into any module from the LMS content menu.
LMS Bridge is licensed under the GNU AGPL-3.0. Any institution can download it, run it on its own infrastructure, and use it with no per-student fee and no size limit.
Clone the repo and bring it up with one command (Docker). It runs inside your environment, not ours — your data never leaves your boundary.
Point it at your own Azure OpenAI / OpenAI / Anthropic key. Student data stays under your contract, and the software costs you nothing to run.
Free for institutions of any size. Optional commercial support, managed hosting, or an OEM/commercial license are available if you want them.
Like the project? Star or sponsor it on GitHub. Need an
SLA, managed hosting, or a commercial/OEM license (to build on it without AGPL's terms)? See
COMMERCIAL.md in the repo or get in touch below.
Clone the repository, copy .env.example to .env, and run
docker compose up. Migrations and a demo seed run automatically.
In the admin console's AI & Privacy tab, point it at your own provider and key — or keep the built-in mock to explore with no external calls.
Register it as an LTI 1.3 tool (see below) and place it in a course. Single sign-on does the rest.
Start with one high-enrollment course; the console reports mastery gains, engagement, and at-risk recovery so you can see the impact.
LMS Bridge registers as a standard LTI 1.3 tool. Your admin needs four URLs from your LMS Bridge instance:
In your LMS admin (Blackboard/Brightspace/Canvas/Moodle) add a new LTI 1.3 / LTI Advantage tool using the four URLs above and enable AGS + NRPS + Deep Linking.
The LMS gives you an Issuer, Client ID, Deployment ID, and its auth/token/JWKS URLs. Enter them in your own admin console (the LMS (LTI) tab) to complete the registration — Canvas and Moodle can do this in one click via Dynamic Registration.
Instructors add LMS Bridge to a module via Deep Linking, and launch — single sign-on does the rest.
One-page PDFs you can hand to a campus LMS admin. The
full combined runbook is in the repo (docs/INSTALL_LTI.md). LMS Bridge targets
1EdTech LTI Advantage certification and a TrustEd Apps listing to ease institutional procurement.
No. Mastery indicators are a private learning aid. Grades are set solely by the instructor through official assessments. Any LMS gradebook write-back is a non-graded column.
The opposite. The tutor is pedagogically constrained: it diagnoses and guides with Socratic questions and never provides graded answers.
LMS Bridge is FERPA-aware with role-based access. For full governance you can run the AI on your own approved infrastructure (e.g. your Azure OpenAI tenant) via bring-your-own-key.
Any LTI 1.3 / Advantage platform — Blackboard, Brightspace, Canvas, Moodle, Schoology, and more — from a single integration.
Yes. LMS Bridge is open source under AGPL-3.0 — free to self-host and use at any size, with no per-student fee. You pay only for your own AI usage (your key) and your hosting. Optional commercial support, managed hosting, and an OEM/commercial license are available if you want them.
GNU AGPL-3.0. You can run, modify, and deploy it freely; if you modify it and offer it as a
network service, AGPL asks you to share your changes. Internal institutional use carries no extra
obligation. See LICENSE and COMMERCIAL.md in the repository.
LMS Bridge is in early pilots — be among the first to put your course on the map. If it's helping your students, we'd love to share your experience (only with your permission).
“Approved quotes from instructors and institutions will appear here.”
— Your words could be first
Add yourself to
ADOPTERS.md via a pull request, or use the form to tell us.
Support the project on GitHub — stars and sponsorships keep it free for everyone.
We never publish your name, quote, or logo without the consent you check above. Logo use also needs your institution's brand approval.
Questions, help installing, an efficacy-study collaboration, or commercial support — send a note and we'll get back to you. (Self-hosting? You don't need us to start — just follow the install guide above.)
Write to contact@lms-bridge.example (replace with your project contact).
Full source is on GitHub
(AGPL-3.0), with an OpenAPI spec at /docs, the LTI provider, and deployment
guides. Issues and pull requests welcome.
A DPA, accessibility statement (VPAT), and security documentation (HECVAT) are available on request.
Built and maintained by Hasan [your surname], New York University. ha2285@nyu.edu · GitHub. (Replace the placeholders with your real name and links.)