Abullahi Yusuf
Founder & CEO
StudyPatch turns short adaptive assessments into clear skill-level insight, then drafts targeted practice for each learning gap. Teachers review, edit and approve everything before a student sees it.
The prototype uses fictional students and data only. No sign-up needed.
| FRC | FOP | INT | OOP | EQN | |
|---|---|---|---|---|---|
| Ines P. | ● | ○ | ○ | ● | ◐ |
| Mateo A. | ● | ◐ | ● | ● | ○ |
| Amara O. | ◐ | ○ | ◐ | ● | ◐ |
Students answer a short adaptive assessment. Questions adjust to their responses within limits you set.
Responses map to curriculum skills. Wrong answers are matched to the reasoning they suggest, such as adding fractions across.
StudyPatch drafts practice, hints and worked examples for each gap. Every item is labelled AI-generated.
You edit, approve and assign. Progress reports show how demonstrated mastery changes over time.
Short, adaptive question paths inside teacher-configured limits: length, difficulty range, follow-ups.
Results mapped to your skill map, showing which concepts need more practice and which are secure.
Practice questions, hints, explanations and worked examples aligned to each gap, checked automatically for a single correct answer.
Class summaries, commonly misunderstood concepts and suggested activities grounded in your own class data.
Track demonstrated mastery across assessments with cautious, evidence-count-aware change indicators.
Teacher, student and administrator experiences with least-privilege access and small-group suppression.
Answer five Year 7 questions, or use “Show example mistakes”. A wrong answer is never just “wrong”: each option is mapped to the reasoning that may sit behind it.
A class heat map and a plain-language summary replace spreadsheet marking.
You set assessment limits, edit generated content, choose who gets it, and can withdraw it at any time.
Draft practice for small groups appears in your review queue, already aligned to the skill map.
Encouraging, plain-language feedback that stays about the work, not about the student.
Progress views and CSV exports using pseudonymous IDs support parent meetings and intervention planning.
Edit skills, objectives and the question bank to match your scheme of work.
Baseline diagnostics at the start of Year 7, then fortnightly check-ins on fractions, integers and algebra readiness.
Short check-ins before a unit to decide which groups need a recap, with practice teachers sign off.
Onboard new learners with a diagnostic, build a first-session plan, and show families progress between sessions.
Aggregated, suppressed school-level views for leaders. Individual answers stay with the teacher.
Identify the specific skills behind a stalled topic and monitor whether targeted practice moves the evidence.
Assign approved practice sets that any teacher can review at a glance.
No AI-generated question or practice reaches a student until a teacher approves it. This control can’t be switched off.
StudyPatch reports evidence on specific skills. It never labels ability or suggests learning disabilities, and it says so on screen.
Every generated question, hint and example is badged “AI-generated”, with the reviewer and date shown to students.
Display name and a pseudonymous ID only. No dates of birth, contact details or health information. Names never go to the AI model.
Built on AWS: KMS encryption at rest, TLS in transit, private networking, role-based access and row-level security.
Administrators set retention windows. Nightly jobs pseudonymise or delete data on schedule, with an audit trail.
Compliance requirements (FERPA, COPPA, UK/EU GDPR and local rules) vary by customer. Design notes are in the documentation and are not legal advice.
Founder & CEO
Co-Founder & CTO
Up to 50 active learners
From 200 students
Multi-school rollout
Illustrative pricing for this prototype. Final pricing is confirmed at demo.
Tell us a little about your school or centre. We’ll reply within one working day. Please don’t include any student information.
No. Generated questions and practice start as drafts. Only content a teacher approves can be assigned, and it carries an AI-generated label.
No. It reports evidence about specific skills from specific questions. It doesn’t make claims about ability, intelligence or learning disabilities. Concerns should go through your school’s usual specialist processes.
Teachers set the minimum and maximum number of questions, the difficulty range, the starting level and whether follow-up questions are allowed. Only approved questions are used, and you can preview example paths before assigning.
A display name, a pseudonymous ID, class membership, and responses to assessments and practice. We don’t collect dates of birth, contact details or health information, and names are never sent to the AI model.
Structured data lives in encrypted PostgreSQL on AWS in your chosen region; educational resources in encrypted S3. Administrators set retention windows and nightly jobs apply them.
Yes. Edit skills and objectives, write your own questions with misconception tags, or approve AI drafts into the bank.
The prototype demonstrates mathematics. The skill-map and misconception model is subject-agnostic and can extend to other subjects.
This website accompanies an interactive prototype built with fictional data. The architecture and rollout plan describe the production path.