The Challenge Addressed
This practice is part of the Export Training Program for Armenian SMEs, a blended programme (March–December 2025) supporting 50 companies through seven export modules and a final export plan.
At the start, three intertwined challenges limited learning impact:
- Low digital skills – Most participants had only basic experience with digital tools; very few could use generative AI meaningfully.
- Strong language barrier (Armenian–English) – Export knowledge, regulations and case studies are mainly available in English, making independent learning difficult between sessions.
- Limited access to export know-how outside Yerevan – Many SMEs lacked exposure to international markets and needed to research standards, partners and markets on their own.
The AI strand was designed to lower digital and language barriers so participants could actively use export tools rather than remain passive listeners.
Why the Practice is Innovative
The innovation lies in how AI was integrated:
- AI as a language bridge and knowledge gateway – Participants learned to use generative AI in Armenian to access global export knowledge, gradually introducing English terminology. AI functioned as a “bilingual tutor” available anytime.
- Embedded in real business tasks – AI was used directly in export activities: market analysis, partner outreach emails, logistics terminology, pricing logic and export plans—always based on the participant’s own company.
- Critical and responsible AI use – Training focused on simple prompting patterns, verification of facts and numbers, and awareness of hallucinations, bias and data protection.
- Low-cost, scalable model – Freely available tools and participants’ own devices (often smartphones) made the approach suitable for development-cooperation contexts.
Type of Integration and Key Competences
AI was layered onto an existing flipped-classroom structure:
- Pre-learning: short AI micro-videos embedded in module introductions.
- Workshops: live prompting demonstrations during export exercises.
- Assignments: each module included an optional or recommended AI step.
- Support: additional hands-on “digital & AI clinics” for less confident learners.
Key competences developed include:
- Digital competence and AI literacy
- Multilingual communication
- Information and data literacy
- Critical thinking
- Entrepreneurial competence
- Learning-to-learn skills
Role of the Teacher/Trainer
Many trainers acted as facilitators and content deliverers:
- Designing export tasks where AI use was meaningful;
- Modelling effective prompting in Armenian and English;
- Guiding safe and ethical AI use;
- Coaching participants with very low digital confidence.
Type of Learning Activities
Typical learning activities included:
- Prompt labs within export modules;
- Language-bridging tasks (Armenian concepts → English export terms);
- AI-supported market research sprints;
- Co-writing and editing of export texts;
- Error-spotting exercises using imperfect AI outputs.
Impact of the Practice
The impact was visible:
- Reduced language barriers and greater engagement with English materials;
- Increased learner autonomy between sessions;
- Higher-quality and more complete export assignments;
- A mindset shift toward continuous, self-directed learning;
- Safer and more critical use of AI.
The practice supports the programme goal that at least 80% of companies take first export steps or enter new markets.
Potential for Replication
The model is highly replicable due to its low cost, tool-agnostic design and modular structure.
It is well suited to adult learning contexts with low digital skills and strong language barriers, and aligns with SDG 4 and EU priorities on digital and entrepreneurial competences.
Innovator: Aigi Young - Visionest Institute OÜ