Request Presentation
Building the early warning layer
universities don't have yet.
Steda is an AI-based system that helps universities detect student disengagement and stress - weeks before academic outcomes change. Concept validated through independent research with 100+ students.
From conversation to early institutional signal.
Steda was tested with real university students to identify early signals of stress, disengagement, and transition risk - before academic outcomes changed.
*Research ongoing — currently expanding to additional university partners.
100+
Students surveyed in independent validation study*
93%
Said an AI mentor would be useful or very useful
96%
Want to try or would consider trying the service
What happens between classes, exams, and reports
Most early student risks emerge outside formal academic checkpoints.
  • Between assessments
    of students cite academic overload as their primary challenge - yet most never report it to their university.
    53%
  • Silent accumulation
    of students turn to official university support when facing difficulties. The rest carry it alone - invisibly.
    1%
  • Late visibility
    of students considered dropping out or taking academic leave - none were visible to their institution.
    15%
By the time academic data changes, the real problem started weeks earlier.
Why this matters institutionally
Early blind spots translate into financial, reputational, and operational risk.
  • Economic impact
    Preventable first-year attrition directly affects tuition revenue and enrollment targets. First-year dropout rates
    of 10-20% are common across universities globally.
  • Rankings & accreditation risk
    Student satisfaction and progression rates increasingly factor into QS, THE, and national ranking methodologies - making early retention a strategic priority.
  • Operational blind spots
    Academic dashboards track grades and attendance - not the stress and disengagement that precede poor outcomes by weeks.
Validated with 100+ students. · 93% find AI mentorship useful. · 96% want to try the service. · Only 1 in 100 used official university support. · Research ongoing - expanding to new universities.
Research Findings
Independent exploratory study — 100+ students, 2026. Expansion to new universities in progress.
  • AI mentorship
    is wanted*
    of students find an AI mentor useful or very useful — demand exists before the product does.

    93%

  • Ready to try*

    want to try or would consider trying the service. Adoption barrier is near zero.

    96%

  • The support gap

    of students turned to official university support when facing difficulties. The system isn't reaching them.

    1%

  • Behaviour
    already exists
    already use AI tools as their primary source of help. Steda meets students where they already are.

    33%

  • Top student
    challenge
    cite academic overload as their biggest difficulty, followed by loss of motivation (39%) and psychological stress (19%).

    53%

  • Silent dropout
    risk
    had considered dropping out or taking academic leave — without any signal reaching their university.

    15%

  • *Based on hypothetical adoption intent survey, n=100+
How It Works
Proposed system architecture — currently in prototype development
Students engage
Students access a 24/7 AI mentor to talk through stress, uncertainty, and academic challenges - anonymously and without judgment.
Signals are aggregated
Patterns are detected across cohorts - no individual tracking, no personal data exposure.
Institutional insight
The university receives early cohort-level signals before academic outcomes change - actionable, structured, decision-ready.
Pilot Model
A Controlled Pilot — Minimal Risk
We are currently seeking first university partners for a no-commitment pilot.
Scope
  • 60-100 students
  • First or second-year cohort
  • One selected program
Duration
  • First signals within 7-15 days
  • 8-10 weeks total
  • No onboarding required from students
What success looks like
  • Clear early-risk patterns identified
  • First cohort-level signals within 15 days
  • Actionable insights for student services
What is NOT required
  • No LMS integration (can be added post-pilot if needed)
  • No grading access
  • No faculty workflow changes
  • No IT infrastructure changes
Why fund this now
Grant support will enable Steda to move from validated concept to first institutional pilot.
Prototype development of the conversational AI agent
01
Institutional dashboard design and validation
02
First pilot deployment in 1–2 partner universities
03
Development of predictive methodology for retention signals
04

Target: First institutional pilot - 2026 

Post-pilot: Subscription per university cohort

From validated concept to scaled product — four focused stages.
1.Concept validated
2026 Q1–Q2
Current stage 
Research complete.
Funding sought.

2.MVP development
2026 Q3
Conversational agent + admin dashboard built.
3.First pilot
2026 Q4
1–2 partner universities
Real cohort data
Retention signals tested.
4.Scale
2027+
Multi-university rollout.
Subscription model per university cohort.
Who's building this
Eugene Maliukou
Product Designer
Maksim Shitik
Full Stack Developer
Daniil Radkevich
AI/ML Engineer
Steda is an early-stage venture. We are a small founding team with deep product and research experience, building in public.
Aliaksandr Bandaryk
Founder, 10+ years in B2B SaaS, EBRD consultant 
Close
Interested in Steda? Leave your contact and we'll reach out with the full presentation and pilot details.
Stay updated on Steda
We share research updates, pilot results, and product progress. No noise.