FlipPraxis UniTech
AI-Powered Personalised Learning for universities that want every student to progress
Each student gets a learning plan built from their own results, with the reason behind every suggestion. Each lecturer gets the evidence and the final say. AI recommends, faculty governs, students demonstrate capability.
- Video · 22 min · RequiredBig-O notationWhy: builds Algorithm Analysis, where your results show a gapEvidence: 34% from 8 answers
- Practice · 6 minPractise Algorithm AnalysisWhy: your lowest result in this courseHigh confidence
- Reading · 15 minStacks and queuesWhy: practises Data Structures at the right level for you
How it works
A continuous cycle of evidence and support
After every assessment the learner profile updates and the recommendations follow.
- 1Sign in through the LMSSingle sign-on identifies the student and their course
- 2Diagnostic checkA short check of prerequisite knowledge
- 3Learner profileMastery estimates and gaps per competency
- 4RecommendationsFaculty-approved lessons, practice and projects
- 5Personalised learningStudy, practise and ask the course assistant
- 6Continuous assessmentAnswers, quizzes and project reviews add evidence
- 7Faculty reviewLecturers approve interventions and resolve exceptions
Try it
See how a recommendation is made
Move the sliders to change a student’s results. The plan is worked out by the same rules UniTech uses, and every step shows why it is there.
- Move a slider to see the plan
For students
A plan, a tutor and projects that fit
- A plan for every course showing what to do next, how long it takes and why
- Practice at the right level, with an explanation after every answer
- A course assistant that explains, gives hints and examples, quizzes them and answers in several languages
- Projects matched to their level and interests, with milestones planned with AI help and verified ProofFolio evidence
- Replies from lecturers to questions the assistant could not answer
For faculty
The evidence, and the final say
- Mastery by student and competency, with the evidence behind every number
- Suggested interventions that lecturers accept, edit or reject
- A question bank and project library where AI drafts stay hidden until approved
- Rubric assessment for individual, group and capstone projects
- Impact, integrity and fairness reports, and AI switches by course, cohort or activity
Mastery levels
What the plan does at each level
Mastery is an estimate from practice, quiz and project evidence, not a grade. Harder questions count for more, and repeating a question the same day does not count twice. Each institution agrees the thresholds with its academic board.
| Mastery | Level | What the plan recommends |
|---|---|---|
| 0–39% | Needs support | Prerequisite lessons and guided practice |
| 40–59% | Developing | Targeted practice and extra examples |
| 60–79% | Proficient | Applied exercises and standard projects |
| 80–100% | Advanced | Enrichment and challenging applications |
Course assistant
Grounded in your own materials
The assistant searches the course’s approved lessons and lecture transcripts, answers from what it finds and cites each source, including the moment in the video. When the materials do not cover a question it says so instead of guessing, and the student can send the question to the lecturer.
- Answers grounded in approved course content, with citations
- Simple, standard or in-depth explanations
- Hints only for graded questions and in exam mode
- Protection against instructions hidden in documents or questions
- Built-in quality testing before students rely on it
Integration
Works with Moodle, Canvas and Blackboard
Students launch UniTech from their LMS through LTI 1.3 without a second login. Rosters sync through LTI or OneRoster, approved results return to the gradebook with faculty approval, and the personalisation core stays independent of any one LMS.
| Connection | Standard |
|---|---|
| Launch and sign-in | LTI 1.3 / OpenID Connect, Microsoft, Google, SAML |
| Courses and rosters | LTI Names and Role Provisioning, OneRoster 1.2 |
| Grades | LTI Assignment and Grade Services, with faculty approval |
| Learning records | xAPI export |
Responsible by design
Safety, privacy and academic integrity
Personalisation only helps if students, lecturers and leadership can trust it. These controls are built into the platform, not added as policy.
- AI never decides grades, discipline or progression
- Students are told when they are using AI and why each activity is recommended
- No names, emails or student numbers are sent to AI models, and no data trains them
- Every recommendation decision, correction and AI use is logged
- A fairness review compares outcomes across student groups
- No learning styles, personality or sensitive attributes are inferred
FAQ
Questions institutions ask
What is AI-powered personalised learning in UniTech?
It gives each student a learning plan built from their own evidence: a short diagnostic check, practice answers, graded quizzes, project reviews and lesson progress. Every recommendation shows the reason, the evidence and a confidence level, and lecturers can change any of them.
Does the AI decide grades or progression?
No. Practice produces a mastery estimate, not a grade. The AI never decides grades, discipline or progression. AI-drafted questions and project ideas stay drafts until a lecturer approves them, and lecturers can correct any estimate with a reason the student can read.
How does the course assistant avoid making things up?
It answers from the course’s approved lesson text and lecture transcripts, cites the lesson and timestamp it used, and says plainly when the materials do not cover a question. Students can send any question to their lecturer with one click.
Can students use the assistant to answer graded work?
The assistant recognises graded quiz questions and switches to hints and concept explanations without giving the answer. Lecturers can also put a course into exam mode, or switch the assistant off for an activity or a cohort.
Which LMS platforms does it work with?
UniTech connects to Moodle, Canvas, Blackboard and other LMS platforms that support LTI 1.3, with single sign-on. A Moodle block puts the course assistant on course pages.
Does UniTech infer learning styles or personality?
No. It does not infer learning styles, intelligence, personality or other sensitive attributes. Personalisation uses assessment evidence, progress and the goals students choose to share.
What student data is sent to AI models?
No names, email addresses or student numbers. Questions and answers are not stored, and they are not used to train AI models. Institutions can switch AI off by institution, course, cohort or activity.
How do we start?
Most institutions begin with one course and 40 to 60 students for a semester through the FlipPraxis Launchpad Pilot: map lessons to competencies, add practice questions and project ideas, then compare diagnostic and later results.
FlipPraxis Launchpad Pilot
Start with one course.
Pilot AI-powered personalised learning with one course and 40 to 60 students for a semester, and measure the change from the diagnostic check.