Production Case Study · Education & Talent Acceleration
Scholarship Hackathon 2026 & Student AI Tutor: Scaling STEM Mentorship in Angola
Programme Highlights: Across our three educational initiatives — the TICNES AI Scholarship Hackathon 2026, the Student AI Tutor, and the TICNES AI BootCamp — we have reached over 1,500 student participants, evaluated 250+ applied AI project submissions, facilitated 10,000+ active tutoring sessions, and trained 500+ students and SME professionals.
Contents
- Addressing the financial and pedagogical barriers in STEM education
- Designing a meritocratic, auditable Scholarship Hackathon
- Engineering the Student AI Tutor: Socratic guidance vs. answer dumping
- Bridging university learning and SME productivity via the AI BootCamp
- Measured educational and community outcomes
1. Addressing the financial and pedagogical barriers in STEM education
University students studying computer science, electrical engineering, mathematics, and applied sciences in Angola face two compounding obstacles. First, annual tuition fees and learning materials represent a severe financial burden for talented students from low- and middle-income families. Second, overcrowded lecture halls and scarce laboratory hours mean that when a student gets stuck on a linear algebra proof, a pointer bug, or a database normalisation exercise at 10 p.m., personalised academic support is unavailable.
At TICNES AI, we view educational platforms not as CSR marketing, but as core systems engineering: building tools that cultivate the next generation of systems builders while funding full university scholarships through transparent, merit-based competition.
2. Designing a meritocratic, auditable Scholarship Hackathon
Traditional scholarship applications often favour candidates who have access to polished recommendation letters or administrative connections. We designed the TICNES AI Scholarship Hackathon 2026 (scholarship-hackathon.ticnes.ai) around a simpler standard: show us a working system that solves a real problem.
- Individual builder format: To ensure every award recipient genuinely authored the work they submitted, the competition evaluates individual builders rather than large uneven teams where one person writes all the code.
- Published multi-criteria rubric: Every submission is scored against four public criteria: Practical Utility for a Local Problem (30%), Technical Execution & Code Quality (30%), Resilience & User Experience (20%), and Documentation & Verification (20%).
- Semi-automated technical triage + human panel authorisation: Automated evaluation pipelines verify repository structure, build reproducibility, and plagiarism/template checks, while final scoring and scholarship award disbursements are reviewed and signed off by named TICNES AI senior engineers.
3. Engineering the Student AI Tutor: Socratic guidance vs. answer dumping
When students use a generic commercial chatbot for homework help, the model immediately prints the final solution or code block. The student copies the output, submits the assignment, and learns nothing — failing the subsequent written university exam.
We engineered the TICNES Student AI Tutor around Socratic pedagogical guardrails:
| Interaction Stage | Generic Chatbot Behaviour | TICNES Student AI Tutor Behaviour |
|---|---|---|
| 1. Problem Submission | Outputs full final solution immediately | Identifies the underlying syllabus concept and asks the student for their first step |
| 2. Intermediate Error | Rewrites the entire answer from scratch | Highlights the specific line where a sign error or logic bug occurred and gives a targeted hint |
| 3. Concept Verification | Ends conversation after printing answer | Generates a short isomorphic practice problem to verify the student mastered the method |
| 4. Bandwidth Footprint | Heavy rich-media SPA requiring fast 4G/Wi-Fi | Lightweight text + KaTeX math rendering optimized for low-bandwidth mobile links |
4. Bridging university learning and SME productivity via the AI BootCamp
Alongside the Student AI Tutor, the TICNES AI BootCamp (ai-bootcamp-training.ticnes.ai) provides intensive, hands-on training for university students, recent graduates, and small-business professionals. Rather than teaching abstract machine-learning history, the curriculum focuses on practical execution: auditing manual administrative workflows, writing deterministic prompt and schema contracts, automating document extraction safely, and measuring productivity gains (detailed in our SME Automation Playbook).
5. Measured educational and community outcomes
- 100% sponsored university tuition scholarships awarded to top individual innovators in the Scholarship Hackathon.
- +88% improvement in concept mastery on post-session diagnostic problems among students using the Socratic step-by-step tutor mode, with a 3.2x improvement in study time efficiency.
- 98.4% satisfaction rate across 500+ students and SME professionals trained through the TICNES AI BootCamp.
To learn more about eligibility for the Scholarship Hackathon or to bring the TICNES AI BootCamp to your university or enterprise, visit scholarship-hackathon.ticnes.ai or contact us.