Learning assurance for the age of AI
Interactive prototype · Illustrative dataAKANYA
Think beyond the answer.
Making genuine learning visible in the age of AI.
Akanya helps higher-education institutions assess how students use evidence, respond to challenge and defend judgement, not only what appears in the final answer.
Interactive functional prototype.
- 1Course evidence
Lecturer-approved sources
- 2Initial position
Recommendation, assumption, risk
- 3Reasoning challenge
One course-grounded tension
- 4Defended conclusion
Defend, modify or replace
- 5Lecturer judgement
Final academic authority
The problem
AI can now produce the final answer. Institutions still need evidence that learning happened.
- 01
Polished written output alone no longer provides sufficient evidence of understanding.
- 02
Detection tools focus on possible authorship rather than whether learning occurred.
- 03
AI tutors may support learning but do not provide institutional assessment assurance.
- 04
Lecturers need better evidence without an unmanageable increase in workload.
Traditional assessment
- 1Question
- 2Final answer
- 3Lecturer review
With Akanya
- 1Evidence
- 2Position
- 3Challenge
- 4Defended judgement
- 5Lecturer review
Lecturer review remains the point of judgement in both models. Akanya gives that review more to work with: the reasoning behind the answer, not only the answer.
Category
Not another AI tutor. A learning-assurance layer.
AI study assistants
Support students through explanations, drafting or dialogue.
Assessment delivery platforms
Deliver, manage and record tests or assignments.
Detection and integrity tools
Analyse final submissions for possible concerns.
Akanya
Capture and review the reasoning journey behind assessed work.
This categoryThe challenge is one mechanism inside Akanya. The product is the complete evidence-to-judgement workflow.
How it works
Five stages, from course evidence to lecturer judgement.
- 1Course evidence
Lecturer-approved sources
- 2Initial position
Recommendation, assumption, risk
- 3Reasoning challenge
One course-grounded tension
- 4Defended conclusion
Defend, modify or replace
- 5Lecturer judgement
Final academic authority
Stage 1
Course-grounded evidence
The lecturer controls the source material and learning outcomes that shape the assessment.
Stage 2
Initial reasoning
The student states a position, supporting evidence, assumptions, risk and rejected alternative.
Stage 3
Reasoning challenge
Akanya identifies one meaningful tension grounded in the approved course material.
Stage 4
Defended judgement
The student defends, modifies or replaces the original conclusion.
Stage 5
Lecturer review
The lecturer reviews the reasoning journey and retains final authority over marks.
Akanya shifts assessment from final-output review to visible reasoning.
Who it is for
Built for the people who design and defend assessment.
Lecturers
- Better evidence of student understanding
- Clearer rubric review
- Visibility of assumptions and misconceptions
Programme and faculty leaders
- Stronger learning assurance
- Responsible AI adoption
- More defensible assessment design
Teaching and learning teams
- Structured assessment redesign
- Course-grounded workflows
- Evidence for pilot evaluation
African in origin, globally relevant
Built for institutions that take assessment seriously.
Akanya is designed for higher-education settings where a defensible view of student reasoning matters — from a single postgraduate module to programme-wide assessment redesign.
- Business schools
- Executive education
- Professional programmes
- Postgraduate faculties
- Teaching & learning units
Prototype
Explore the working prototype.
The guided experience introduces the Akanya assessment journey. Approved academic reviewers may also request access to the separate interactive prototype currently under development.
Prototype access is reviewed before an invitation is issued.
Responsible AI
Human academic authority remains central.
Akanya provides decision support. It is designed so that judgement stays with the academic, and so that students can see and respond to the evidence behind any decision.
- No automatic misconduct decisions
- No automated final marks
- No behavioural or typing-profile analysis
- Lecturer-approved course material
- Transparent source references
- Student opportunity to respond
- Minimum necessary data
- No public-model training on student work
- Accessible alternatives
- Institution-controlled policy settings
Pilot
Seeking one controlled institutional pilot.
Suggested scope
- One module
- One assessment
- 100–250 students
- One academic term
- Lecturer-controlled marks
- Lightweight implementation
- Pilot evaluation report
Phase 1
Co-design
Align the assessment, sources, rubric and AI-use policy.
Phase 2
Controlled assessment
Run one defined assessment under lecturer oversight.
Phase 3
Pilot evaluation
Review academic usefulness, student experience, lecturer workload and governance requirements.
When AI can produce the answer, institutions need a better way to see how students think.
Akanya is a Teddy K Group venture.