The AI System Helping Schools Spot Student Struggles Before They Turn Into Crises

Article written by Bailey Shaltner  Date 08-29-25

Violence in schools is not a new issue, however, it is one that receives a lot of media attention due to the frequency of headlines making it into the news cycle. The United States already averages over 30 school shootings each year, and in 2025 we are over halfway through the year with already 30+ school shootings1 (CNN, 2024)

 

Just this month, a shooting in Minneapolis occurred where two children were killed and 18 others were injured(The Guardian, 2025). What is even more alarming, since 1999, the percentage of students who have experienced gun violence has almost tripled and has lifelong implications on students’ mental health, school outcomes, and general trust in their community3 (KFF, 2025).

 

But not every crisis makes national headlines. For most schools, the warning signs begin much earlier and in much quieter ways – a student pulling back from friends, falling behind on assignments, or going silent in class. These are the moments when intervention matters most.

A proactive answer

However, there is a disconnect when we wait for these mass school shootings to occur before we feel the need to do something – something is already too late. The way to prevent these large-scale incidents is to pay attention to those around us, offer support, and connect with students before it is too late, rather than using zero tolerance policies and labeling students with criminal records for life.

Mariya Giy, creator of the Early Insight & Engagement System (EIES), is fusing science and software to help schools notice early signs of student struggle – so adults can step in with support before a crisis. Co-created with Dr. Anastasiia Timmer of California State University, Northridge (CSUN), powered by Hackico.AI, EIES is grounded in research and built for the cadence of school life.

Spotting subtle shifts

EIES doesn’t diagnose or label students. When a student quietly shifts from the front row to the back, turns work in right at the bell, or goes silent in small-group discussions, the signal is subtle – and easy to miss. By the time it’s unmistakable, momentum is already gone. EIES was created to move that moment earlier.

“The premise is simple: when schools notice small changes early, they can respond with care,” says Mariya Giy. “EIES puts that idea into practice with plain-language insights teachers can actually use.”

Why schools need a different kind of AI

Educators are dealing with an unprecedented mismatch between student needs and time/resources. A recent survey found that schools have seen material increases in student anxiety, absenteeism, and behavior, as well as teacher burnout and attrition since 2020(RAND, 2023). They are also operating with extremely full plates managing rosters, assignments, family communications, and planning across numerous tools that don’t communicate or share information. Current school-facing analytics are focused on lagging and often siloed systems (static dashboards or “what happened” or crude risk flags that can stigmatize students and overlook meaningful change).

 

EIES was designed differently. It’s built on the low-stakes, high-volume signals teachers and leaders see every day: small shifts in participation and engagement, subtle patterns in teacher-student interactions, peer relationships or group dynamics, family interactions, and more. Evidence from the University of Chicago’s Consortium on School Research and other research centers shows that small changes in classroom participation or peer connection are the leading indicators of more disruptive academic or behavior challenges later on5 (University of Chicago Consortium on School Research, n.d.).

Focusing on these fine-grained, non-stigmatizing signals, EIES is helping teachers and leaders identify these inflection points early enough that they can respond with information and clarity, not crisis.

From science to software

Mariya Giy’s route to AI began with an M.A. in data-centric Empirical Sociology, studying how social context shapes behavior and belonging. That lens – ask what changed, for whom, and in what environment? – eventually met large-scale engineering in education. The crossover is where EIES was born: a system designed not to assign labels but to prompt sooner conversations.
With CSUN professor collaboration, the team translated research on engagement and school climate into product requirements: prioritize change over time, keep outputs short and actionable, and ensure every prompt can be explained in plain language.

“We wanted something teachers could trust on Friday afternoon,” Dr. Timmer notes. “A brief that says: here’s what shifted, here’s why it might matter, here are a few ways to respond on Monday.”

What EIES is (and isn’t)

What it is: an early-insight and engagement layer that sits alongside the systems schools already use and produces a concise weekly brief for teachers, counselors, and families.

What it isn’t: a clinical diagnostic; a behavior score; a tool that replaces professional judgment.

By focusing on drift – small but sustained changes over time – EIES keeps the door open to context: a new job at home, sports season, transportation hiccups, or the ordinary turbulence of adolescence. The brief invites curiosity first.

 

A Friday brief, in practice

Imagine a teacher opens Friday’s EIES summary. No alarms. No tiers. Just a pattern of change:

  • Two assignments submitted right at the bell (previously early).
  • Fewer comments during a small-group lab.
    Caregiver replies slowed from weekly to bi-weekly.

Suggested actions:

  • A two-minute check-in about workload and interests.
  • Pair the student to co-lead next week’s lab.
  • Send a short “how’s it going?” note home with two flexible times for a call.

None of this requires a new program or extra paperwork. It simply makes the human move: the conversation, more likely to happen.

Built for the cadence of school life

EIES runs on a weekly rhythm: fast enough to catch drift before it calcifies, measured enough to avoid notification fatigue. Briefs are short, written in plain language, and aligned with real planning moments – Friday reflections, Monday prep, grade-level meetings.

Behind the scenes, a modular pipeline securely connects to approved data sources (attendance markers, participation signals, communication logs). Because the architecture is component-based, districts can start small – with the minimum inputs they’re comfortable using – and add more later without disruption. The aim is practical: fit the tool to local practice, not force a one-size-fits-all playbook.

Guardrails by design

Putting AI near students demands clarity, control, and restraint. EIES builds these in from the start:

  • Minimal data, maximum clarity. The system is frugal with data and explicit about what it uses. Every prompt includes a plain-language rationale: what changed and why it might matter.
  • Trends over traits. EIES highlights change over time rather than tagging students with fixed labels, reducing stigma and preserving nuance.
  • Human in the loop. Suggestions are just that—suggestions. Educators can accept, modify, or dismiss them. The final decision stays human.
  • Transparency and control. Administrators configure sources, visibility, retention, and escalation paths; families deserve clarity too. The framing is support, not scorekeeping.
  • Bias safeguards. Design choices (no risk tiers, explainable prompts) and governance (auditable settings, periodic reviews) help identify and mitigate blind spots.

“Technology should strengthen relationships, not replace them. EIES succeeds when it makes the human outcome visible.” — Mariya Giy

Measuring impact without over-measuring students

Schools don’t need another dashboard; they need evidence that time is better spent. EIES proposes human-centered indicators:

  • Did check-ins happen earlier than they otherwise would have?
  • After small adjustments (flexible deadlines, peer pairings, quick conferences), did engagement improve the following week?
  • Are teachers reporting fewer surprise escalations because subtle dips were noticed sooner?
  • Do families feel more connected to what’s happening at school?

None of these require tracking students more aggressively. They ask whether timely attention changed the next seven days.

Implementation without disruption

Adoption rises when tools meet educators where they already work. EIES delivers its weekly brief through existing channels, email summaries, staff portals, or embedded in planning workflows, so there’s no new tab demanding constant checks. Rollouts begin with secure connectors to approved sources, simple configurations for which signals to monitor, and a short onboarding: how to read a summary, how to log “we checked in,” and how to tune suggestions over time.

Why now – and why this approach

Talk to teachers and counselors anywhere and you hear the same tension: needs are up, time is down. The temptation is to search for a silver-bullet model that predicts everything. EIES resists that. Instead, it treats insight as a lightweight catalyst: a two-minute conversation that might not have happened, a small flexibility that keeps a student engaged, a prompt that reconnects home and school before silence stretches too long.

 

Beyond EIES: Hackico.AI

Mariya Giy is the founder and CEO of Hackico.AI, an R&D organization at the nexus of AI, EdTech, applied research, and behavioral science. A multi-disciplinary team of engineers, researchers, and educators powers EIES’ technology engine and works to make humanistic and at times abstract AI solutions practical by developing products and platforms. Focused on the research, and development of explainable AI to improve educator decision-making, student well-being, and timely, data-driven interventions, Hackico.AI is changing how schools are able to detect and act on students’ needs before they start to fall behind.

 

“As an experienced Data Analyst, I know how powerful real-time data analytics can be. EIES and Giy’s leadership offer the opportunity to make AI proactive and strategic for schools. I am excited to be in a position where we are using advanced data insights to recognize risk early and enable effective and timely interventions that have a meaningful impact.”

Oleksandr Rovnyak, Data Analyst at Hackico.ai, former Team Lead IKEA

 

EIES is powered by Hackico.AI. The organization collaborates with academic researchers (including CSUN) and school teams to keep solutions practical, explainable, and in line with the weekly rhythm of K–12. (Company site: https://hackico.ai). Under Mariya Giy’s leadership, Hackico.AI is defining a new category of responsive, research-informed AI for K–12: intelligent systems that adapt to the needs of students, families, and the educators supporting them.

Looking ahead

The near-term roadmap focuses on school-friendly pilots with clear defaults: start small, measure what matters, and let teams tailor prompts they find most helpful. Next modules will explore adjacent use cases where early context helps: easing re-entry after extended absences, smoothing transitions between grades, and supporting multilingual families with better timing and clarity.
In every case, one bar remains: does this make it easier for caring adults to do the right thing this week? If schools can catch even a fraction of these subtle shifts earlier, we create not just safer schools, but more human ones – where students are seen and supported before struggle becomes crisis.

References:

  1. CNN. (2024). School shootings fast facts. Retrieved from https://www.cnn.com/us/school-shootings-fast-facts-dg/index.html
  2. The Guardian. (2025, August 27). Minneapolis school shooting leaves two children dead and 18 injured. Retrieved from https://www.theguardian.com/us-news/2025/aug/27/minneapolis-school-shooting-shock-horror
  3. KFF (Kaiser Family Foundation). (2025). Examining school shootings at the national and state level and mental health implications. Retrieved from https://www.kff.org/mental-health/issue-brief/examining-school-shootings-at-the-national-and-state-level-and-mental-health-implicationsʼ
  4. RAND Corporation. (2023). Educator perspectives on student well-being and staffing challenges in U.S. public schools. Retrieved from https://www.rand.org/pubs/research_reports/RRA1108-8.html
  5. University of Chicago Consortium on School Research. (n.d.). Foundations for Young Adult Success: A Developmental Framework. Retrieved from https://consortium.uchicago.edu/publications/foundations-young-adult-success-developmental-framework