Please ensure Javascript is enabled for purposes of website accessibility
About FERIT
GeneralNewsDepartmentsStaffLiving in OsijekHow to reach OsijekUseful websites and accomodationDocuments
Study Programmes
GeneralUndergraduate University StudyGraduate University StudyUndergraduate Professional StudyAutoComCourses in EnglishPostgraduate
International Cooperation
GeneralIncoming staff mobilityIncoming student mobilityBlended intensive programmesAILEM conferencePartner universitiesIndustryMikroTik Academy
Research & Development
GeneralResearch GroupsProjectsConferencesJournal
Hrvatski / CroatianPrijava

 Blended intensive programmes

 

 

SUMMARY

This blended intensive programme introduces participants to artificial intelligence and, in particular, to AI agents: systems that do not only answer questions but act on a user's behalf by using tools, files and online services. Designed for students, academic and non-academic staff with no programming background, the course combines conceptual foundations with hands-on work on participants' own everyday study, teaching and administrative tasks.

The program begins with an overview of how AI works, from its history and the basics of machine learning to large language models, and explains what distinguishes an agent from a chatbot. Participants then focus on the language of delegation: writing clear task instructions, examining how wording and language choice affect results, and applying these skills to tasks such as email triage, meeting follow-up and weekly planning.

A central theme of the course is human oversight. Participants learn how AI output fails, practice detecting errors and fabricated sources, discuss academic integrity and disclosure, and examine personalization, privacy, security and the legal framework governing AI in Europe, including the GDPR and the EU AI Act. The final day addresses user-centered design of AI tools and the ethical and societal impact of AI.

Throughout the week, mixed international teams develop a capstone project: a test of an AI agent on real everyday tasks, presented together with the instructions given, the checks applied to its output and the team's conclusions on where it can and cannot help. Regular feedback ensures participants leave able to use AI agents effectively, critically and responsibly.

 

 

GENERAL INFORMATION   |  AGENDA  |  BROCHURE  |  LEARNING OBJECTIVES / OUTCOMES  |  COURSE CONTENT  |  METHODS AND ASSESSMENT

BIP TITLE AI for everyday tasks
BIP code 2026-1-HR01-KA131-HED-000415038-1
Lead Instructor / BIP Coordinator

Assistant Professor Dragana Božić Lenard, PhD
e-mail: dragana.bozic@ferit.hr


Assistant Professor Hrvoje Leventić, PhD

e-mail: hrvoje.leventic@ferit.hr

Contact hours 40 face-to-face + 6 online (90 hours total workload)
Dates for physical component May 24-28, 2027
Dates for virtual component

May 4 and May 18,2027 (pre-phase)

June 11, 2027 (post-phase)

ECTS credits 3
Target audience Undergraduate and graduate students, (non)academic staff
Prerequisites English language B1-B2 proficiency level; no programming experience required
Websites for information  https://www.ferit.unios.hr/international-cooperation/blended-intensive-programmes 
 
Application procedure APPLY / ENTER APPLICATION FORM

Host University: 
Josip Juraj Strossmayer University of Osijek – HR OSIJEK01

Host Faculty: 
Faculty of Electrical Engineering, Computer Science and IT Osijek - FERIT

Partner Universities:  
"1 Decembrie 1918" University of Alba Iulia, Romania, RO ALBAIU01
Technical University of Varna, Bulgaria, BG VARNA02

Practical info:   
Student center school   Bus  directions_bus  Train train  Social events  group

 

Understand how AI and AI agents work

  • explain basic concepts of artificial intelligence, machine learning and large language models
  • distinguish between chatbots and agents in terms of tools, memory, planning and permissions

Delegate everyday tasks to AI agents

  • identify recurring tasks that are suitable and unsuitable for delegation to AI
  • write clear task instructions and design a multi-step agent workflow for a real task

Use language strategically when working with AI

  • analyze how wording, register and ambiguity affect AI output
  • compare AI performance across English, other languages and mixed-language instructions

Evaluate and verify AI output

  • recognize typical AI errors such as fabricated references, omissions and overconfident summaries
  • apply verification routines before relying on or sharing AI output

Use AI responsibly, safely and lawfully

  • interpret institutional AI policies and disclose AI use appropriately
  • assess privacy and security risks of connected agents and apply GDPR and EU AI Act principles

Reflect on AI in people’s lives and society

  • evaluate AI tools for usability, accessibility and transparency
  • discuss the ethical and societal impact of AI and develop a personal plan for its use

ONLINE

Online session 1: May 4, 2027

 

Task diary: May 4-11, 2027

 

Online session 2: May 18, 2027

 

Post-phase reflective report: deadline Friday 11 June 11, 2027

Virtual component

 

  • Online session 1: introductions, mixed-team icebreaker and pre-survey on participants’ current use of AI.
  • Task diary: a one-week log of recurring tasks (emails, scheduling, reading, notes, forms, reporting) that participants revisit at the end of the course.
  • Online session 2: setting up the tools and a first demonstration of an AI agent completing a simple task.
  • Post-phase: individual reflective report (1,500 words or a short video) on one task the participant now delegates to AI and how they supervise it.

DAY 1

May 24, 2027

How AI works: from rules to agents

 

  • Welcome and orientation: brief overview of the course goals, schedule and expected outcomes.
  • Mini-lecture: a short history of AI, from rule-based systems to machine learning, large language models and agents.
  • Workshop: how machines learn (training data, models and errors).
  • Mini-lecture: inside a language model: tokens, next-word prediction, the context window and why a model "forgets".
  • Demonstration: how language models generate text and why fluent output can still be false.
  • Session: from chatbot to agent: learning from feedback, tools, memory, planning and permissions.
  • Demonstration: an agent working through a prepared everyday task, explained step by step.

DAY 2

May 25, 2027

Talking to AI: the language of delegation

 

  • Mini-lecture: AI as a language machine: ambiguity, register and implicit assumptions in instructions.
  • Workshop: writing a task brief (role, context, constraints, output format and what counts as “done”).
  • Activity: rewriting weak instructions into strong ones and comparing the outputs.
  • Experiment: giving the same task in English, in participants’ own languages and in a mix of languages; participants judge the accuracy, naturalness and register of the output as native speakers.
  • Lab: everyday tasks such as email triage, turning meeting notes into action items and weekly planning, with role-specific tasks for students, academic and non-academic staff.
  • Peer review: participants exchange and critique each other’s task briefs.

DAY 3

May 26, 2027

Trust but verify

 

  • Mini-lecture: how AI fails: fabricated references, omissions, agreeing with the user, overconfident summaries and agents reporting unfinished tasks as complete.
  • Challenge: error-detection lab in which teams find and classify errors in AI-generated texts.
  • Activity: “Human or AI?” using the LABUS app, followed by a discussion of the limits of AI detectors.
  • Session: academic integrity and disclosure: comparison of institutional AI policies.
  • Activity: drafting a personal AI-use statement for one’s own studies or job.

DAY 4

May 27, 2027

Agents, personalization and the law

 

  • Session: memory and personalization in assistants: what an assistant stores about a user, where it is kept and how it shapes later answers..
  • Activity: auditing the memory and personalization settings of participants’ own AI tools.
  • Demonstration: privacy and security risks of connected agents, including hidden instructions that hijack an agent (prompt injection).
  • Activity: drafting a permissions checklist for connecting agents to email, files and accounts.
  • Mini-lecture: legal aspects of AI: GDPR, basics of the EU AI Act, and copyright and ownership of AI output, illustrated with everyday cases.
  • Capstone kick-off: mixed teams choose a real everyday task and start building a multi-step agent workflow.

DAY 5

May 28, 2027

AI, people and society

 

  • Workshop: designing AI for people: teams evaluate two AI interfaces for clarity, accessibility and how they communicate uncertainty.
  • Debate: ethics and societal impact of AI (bias, jobs and education, environmental cost, over-reliance) with assigned positions.
  • Capstone presentations: what the agent does, how it was instructed, how its output is checked and what it is not allowed to do.
  • Feedback: instructor and peer evaluation of the capstone projects.
  • Reflection and closing: post-survey, personal action plan and certificates.
METHODS

Interactive workshops

Hands-on sessions to introduce key concepts and techniques, allowing participants to learn through guided practice with AI tools.

Mini-lectures

Concise presentations providing theoretical grounding on topics such as machine learning, language models, AI failure modes and the legal framework.

Live demonstrations

Instructor-led demonstrations of how AI tools and agents behave, including how they fail and how they can be misused.

Hands-on labs

Guided practice on participants’ own everyday tasks, differentiated for students, academic and non-academic staff.

Task diary

A log of recurring tasks kept during the virtual phase, providing authentic material for activities throughout the course.

Comparative experiments

Structured comparisons of AI output across instructions, languages and tools to show how input shapes results.

Error-detection challenges

Team-based tasks in which participants find and classify errors in AI-generated texts.

Case studies

Short everyday scenarios on privacy, copyright and academic integrity to analyze and resolve in groups.

Structured debate

Discussion with assigned positions on the ethical and societal impact of AI, fostering critical thinking and argumentation.

Collaborative capstone project

Mixed international teams design, test and present an agent workflow for a real everyday task.

Peer review

Participants exchange task briefs and project work and provide constructive feedback.

Reflection

Guided discussions and a reflective report connecting course insights to participants’ own study and work.




ASSESSMENT 
 

Participation and engagement

Evaluation of participants’ active involvement in workshops, labs and discussions.

Observation checklists to track engagement and collaboration in mixed teams.

Portfolio assessment

Assessment of tasks completed during the course, such as the task diary, task briefs, error-detection analyses, the permissions checklist and the personal AI-use statement.

Instructor’s feedback

Ongoing, formative feedback after activities, focusing on strengths and areas for improvement.

Capstone presentation

Team presentation of an agent workflow for a real everyday task, focusing on instructions, verification of output and limits on the agent’s actions.

Instructor’s feedback and peer evaluations.

Reflective report

Individual report (1,500 words or a short video) submitted in the virtual post-phase on how the participant delegates and supervises one everyday task.

Application form

Registration deadline has elapsed!

Documents

Blended intensive programmes - Documents:

Kneza Trpimira 2B, HR-31000 Osijek | Cara Hadrijana 10b, HR-31000 Osijek Tel: +385 (0) 31 224-600 | Fax: +385 (0) 31 224-605

IBAN: HR19 2390 0011 1000 16777, HPB | OIB: 95494259952 | PDV id. / VAT id.: HR95494259952 © 2026 FERIT | ferit@ferit.hr