EVA
Educational Virtual Assistant
EVA is your Educational Virtual Assistant
Faculty-validated knowledge.
Professor-defined assistance.
Demonstrable learning.
EVA is MyFaculty.AI’s governed digital teaching presence—the digital counterpart of a course teaching assistant.
Grounded exclusively in faculty-validated course materials, EVA provides learners with real-time, multilingual support via phone, mobile, and computer while preserving academic control, institutional ownership, and IT governance.
Professors define the assistance EVA may provide for each course, learning activity, and assessment, ensuring that it scaffolds learners’ reasoning rather than replacing it—and makes learning demonstrable, not merely answers deliverable.
Dedicated dashboards give faculty and IT teams oversight of teaching activity, quality, usage, and operations.
A Digital Teaching Presence for Every Course
EVA is a real-time, voice-enabled digital teaching presence deployed per course, per semester. She is not a replacement for the professor: she is the digital counterpart of a course teaching assistant, grounded exclusively in the faculty-validated materials of a specific course — syllabus, readings, lecture recordings, transcripts, and assessments.
Learners can study with EVA at any moment, on any day, in their own language: reviewing a concept, preparing an assessment, deep-diving a chapter, or testing their knowledge against the course’s learning objectives.
EVA is multilingual: the course content stays the professor’s, while the conversation adapts to the learner.
EVA is live today — schedule a demonstration and you will see her teaching a real course, and can question her yourself.
Governed, not improvised.
At the core of EVA is the Teaching Interaction Model (TIM), MyFaculty.AI’s proprietary instructional-intelligence layer. TIM determines how EVA teaches — when to explain, question, hint, guide, or redirect — and how much assistance she may provide in each learning situation.
The professor defines the boundaries.
TIM applies them throughout the conversation.
What “governed” means?
EVA does not operate as an unconstrained AI assistant. Her teaching behavior, knowledge, data use, and system operation are controlled within defined academic and institutional boundaries. Faculty govern the validated course knowledge, TIM governs how EVA teaches, ETHOS supervises responsible AI behavior, and ORION monitors operations, security, and system health. Professors and IT teams retain visibility and oversight through dedicated dashboards.
EVA operates within a closed, faculty-validated course knowledge base, cites the sources behind every answer, and redirects questions outside her scope back to the professor.
Autonomous agents supervise pedagogical integrity, security, and system health.
The Architecture Behind Every Conversation
Behind every learning conversation is a coordinated system of specialised agents — governed, not improvised.
TIM — Teaching Interaction Model. EVA’s instructional-intelligence core. It combines the course knowledge base, an orchestrator that coordinates each teaching interaction, and retrieval intelligence that finds and connects the right course material behind every answer.
ETHOS — Responsible AI for Education. The trust, ethics and data-governance layer, ensuring every interaction remains ethical, safe, and aligned with institutional policy.
DELPHI — Learner Analytics. Turns learning conversations into meaningful insights, feeding the Professor Dashboard with the real learning demand of each cohort.
ORION — Operations Control Center. Keeps the platform healthy, secure, and performing — serving the IT team’s supervision, EVA quality monitoring, and cybersecurity and compliance.
What learners experience as one consistent teaching presence is, underneath, an institutional system — with every layer visible and accountable to the professor and the IT department.
TIM is the core Teaching Interaction Model built on a Specialized Reasoning Model
TIM is a governed pedagogical intelligence system that uses a specialized reasoning model to decide how EVA should teach.
TIM determines how EVA should teach.
The foundation model provides language and reasoning capabilities.
Governed knowledge defines what EVA may teach.
EVA on the Mobile
Quick learning support for students on the move, ideal for short revision sessions, concept clarification, and continued learning outside class.
EVA by Phone
Call EVA from a traditional phone — no app or internet connection required. Learners identify themselves with their personal PIN code and can immediately begin a voice study session. Phone Access extends academic support to situations where a smartphone, computer, or data connection is unavailable.
EVA on the Computer
A richer study environment with written explanations, visual support, spoken tutoring, and session history. EVA Text Chat supports structured written interaction; EVA Live Tutor combines spoken explanations with EVA View, captions, microphone, and the EVA avatar; and Past Sessions lets learners revisit earlier questions and explanations.
What You Can Do with EVA
- Ask EVA General Questions about the Course
- Get Ready for the Assessments
- Deep Dive on a Book Chapter
- Review your Knowledge
- Practice with a Quiz
- Get a Study Guide
- Review your Progress
The Right Support for Every Learning Moment
EVA does not provide the same kind of assistance in every situation. Her support is configured according to the course, the learning activity, and the assessment rules.
Through five pedagogical modes, EVA provides the appropriate level of guidance while preserving learner agency, cognitive effort, and academic integrity.
Learn
EVA explains concepts, asks questions, connects ideas, and checks understanding using the professor’s validated course materials.
Practise
EVA challenges learners through questions, simulations, quizzes, and progressively revealed hints—helping them reach an answer without removing the thinking required to learn.
Produce
EVA supports brainstorming, structure, critique, and revision within defined rules. The learner remains responsible for the ideas, analysis, judgement, and final work.
Assess
EVA recognises protected assessment situations and respects the assistance boundaries defined by the professor. She can clarify permitted instructions while refusing support that would compromise academic integrity.
Five pedagogical modes
One faculty-controlled teaching presence
Reflect
EVA helps learners revisit their reasoning, identify knowledge gaps, connect feedback to learning objectives, and determine what they should explore next.
Grounded and Institution-Controlled
EVA answers exclusively from the faculty-validated course knowledge base.
She does not retrieve information from the internet when answering, cites the sources behind every answer, and openly acknowledges when a question falls outside her scope — directing the learner back to the professor.
Course content remains the intellectual property of the institution. The knowledge base is non-transferable and can be hosted on GDPR-compliant infrastructure in the EU, on-premises, or in-country to meet national requirements, with sovereign AI options available.
Teaching Insights Between Classes
Professor Dashboard
For faculty, the Professor Dashboard (Teaching Insights) reveals what students actually struggle with: most-asked questions, most challenging topics and assessments, engagement trends, and each learner’s progress across the semester.
EVA gives professors visibility into the learning happening between classes, allowing them to prepare targeted explanations, adjust classroom discussion, and identify learners who may need additional support.
Across the semester, learning interactions create a documented trajectory of questions asked, concepts revisited, and progress against learning objectives — providing evidence of the learning process, not only the final submitted artifact.
Demonstrable Learning
EVA helps make the learning process visible—not only the final submitted answer. Across the semester, learning interactions can create a documented pathway of questions asked, misconceptions addressed, concepts revisited, feedback considered, and progress made against learning objectives.
This provides professors with evidence of how learning develops between classes while helping learners reflect on their own progress.
The objective is not to surveil learners or determine whether a text was written by AI. It is to answer a more meaningful educational question:
Can the learner demonstrate what they have understood, how their thinking developed, and what remains to be learned?
Where appropriate and subject to institutional policy, consent, privacy, and retention rules, these learning pathways can provide:
- Evidence of engagement with course concepts.
- Visibility into recurring misconceptions and knowledge gaps.
- Traceability of the guidance and sources provided by EVA.
- Indicators of progress against learning objectives.
- Material for learner reflection and faculty intervention.
Scaffolding, not answer-giving
EVA supports the learner’s reasoning without replacing the cognitive effort required to learn.
Evidence of learning, not AI detection
EVA helps make the learning process visible through questions, reflection, source use, and documented progression.
IT Operations Center
Enterprise-Grade Visibility and Control
For the university’s IT team, the IT Dashboard (EVA Operations Center) provides full technical and operational supervision of the platform.
For IT teams and administrators, the EVA Operations Center provides technical, operational, and cybersecurity supervision of the platform. It gives the institution continuous visibility into activity, knowledge-base health, users, infrastructure, security, and backend performance.
It covers four areas:
- Channels & activity — sessions, user/EVA turns, latency, and recent records across EVA’s Phone, Mobile, and Computer access channels and their associated interaction modes.
- Knowledge base — confirm that EVA is grounded in the correct sources; identify weak retrievals, no-hit records, and missing source metadata.
- Users, courses & calendars — review assigned and active users, login sessions, password-change flags, and teaching-week boundaries.
- Operations & cybersecurity — monitor CPU, RAM, disk, API usage, error records, platform health, security events, cybersecurity signals, and backend-health conditions.
The IT department retains the visibility and control it expects from an enterprise system, including operational health and cybersecurity monitoring — no black box.