Threshold: Knowledge-Graph-Driven
Academic Advising Assistant

Scale authentic advising from semester check-ins to daily, proactive support — helping advisors understand why a student needs attention, not just that they do.
Bobby Hodgkinson Smead Aerospace Engineering, CU Boulder  ·  Ryan St. Pierre Threshold Labs
Threshold
Student Success
Conference 2026
Advisors need context,
not just alerts
  • Complex pathways — nonlinear degree paths with minors, certificates, and interdisciplinary interests
  • Fragmented data — engagement, grades, and catalog info in separate silos
  • Reactive advising — meaningful conversations happen once or twice a semester, not daily when students actually need them
  • No root-cause visibility — prerequisite gaps and workload spikes invisible until too late
Early warning
with explanations

The dashboard delivers daily proactive alerts — surfacing who needs attention and why. Instead of reactive semester reviews, advisors get continuous intelligence: risk factors, prerequisite gaps, engagement shifts, and strengths-based opportunities updated every day.

Every day, for every student:

  • Why they're flagged, not just that they are
  • Strengths and opportunities, not just deficits
  • Life context that shapes academic performance
  • Connected advisor team and intervention history
Threshold
Advising · CU Boulder
CU

Academic Advising Dashboard

Monitor student progress and coordinate interventions

Last synced: 2 hours ago Sync Canvas Data
Critical Alerts
Requires immediate attention
5
Early Warnings
Showing signs of struggle
1
Intervention Success
Improved after intervention
78%
Total Students
In your advising cohort
14
Risk Distribution 94% retention rate
Critical (5) Warning (1) Watch (3) On Track (5)
MT
Marcus Thompson critical
Computer Science · Sophomore
Failing Data Structures with grade below... Significant GPA drop from cumulative...
1.40
Semester GPA
ER
Emma Rodriguez critical
Psychology · Junior
Minimal Canvas activity — logged in only... Sudden disengagement from previously...
1.80
Semester GPA
JM
Jake Morrison critical
Aerospace Engineering · Junior
Failing Aerodynamics (ASEN 3111) — core... Thermodynamics below passing...
1.90
Semester GPA
The system surfaces why, not just who
Prerequisite Gap
“Struggling in Algorithms — weak Discrete Math grade. Prerequisite gap likely contributing.”
Engagement Shift
“3 late assignments after perfect history — sudden disengagement, check-in recommended.”
Strengths Match
“Strong ML-adjacent performance — Data Science minor only 2 courses away.”
Workload Alert
“18 credits + new part-time job — may exceed sustainable threshold.”
Three roles,
one shared graph
🎓
Students Navigate complex degree paths, discover minors and certificates aligned with their interests, see prerequisite chains before enrollment
👥
Academic Advisors Scale authentic connection across 300–500 students — daily proactive alerts replace once-a-semester check-ins with continuous, contextual support
📚
Faculty & Administrators Update curriculum to match university goals and market-ready outcomes — see where prerequisite gaps cause downstream failures
MT

Marcus Thompson

Computer Science / Data Science · Sophomore · ID: 108234567

×
1.40
Semester GPA
2.80
Cumulative GPA
15
Credits
100
Risk Score
✉ Email Student
📅 Schedule Meeting
Overview
Courses
Interventions
History
🎓 Academic Advisory Team
Sofi Alarcon Escallada Primary Advisor
Computer Science · ECOT 717
Cole Gladhart Minor Advisor
Mathematics · Math 240
📚 Life Context & Key Events
💼 Employment
25 hrs/week
The Sink Restaurant · off-campus
🏠 Housing
Off Campus
15 min commute
From fragmented data
to connected context

Threshold builds a knowledge graph from existing university data sources. Each data point gains meaning from its connections to everything else. Connects via standard Canvas LTI integration — no custom infrastructure required.

Canvas LMS (LTI) Transcripts Course Catalog Advising Notes
Threshold CU Boulder Course Catalog — Knowledge Graph View
Organic Chemistry
History
Communications
Sociology
Music
Philosophy
Discrete Math
Structural
Linear Algebra
Biochemistry
Optimization
Networks
Algorithms
Computer Vision
Programming
Machine Learning
Databases
Robotics
NLP
Data Science
Psychology
HCI
Neuroscience
Entrepreneurship
Systems
Open questions
  • Canvas integration — what engagement signals matter most for early intervention?
  • Advisor workflow — how does contextual intelligence change the advising conversation?
  • Curriculum feedback — can prerequisite gap data help faculty identify structural issues?
  • Equity & access — does consistent early warning reduce intervention disparities?
See it in action
Scan to explore the live demo, learn about our Canvas LTI integration, or talk with us about bringing Threshold to your institution.
thresholdlabs.io/advising
getthreshold.com
What this looks like
on campus
  • 🎓
    Student: “What can I minor in?” Graph shows a CS student is 2 courses from a Data Science minor based on completed coursework and topic overlap
  • 👥
    Advisor: 500+ student caseload Daily proactive alerts surface which students need attention today — turning once-a-semester check-ins into everyday contextual support
  • 📚
    Faculty: curriculum review Prerequisite chain analysis shows that 40% of Algorithms failures trace back to weak Discrete Math foundations
  • 📝
    Accommodations: exam scheduling Student needs extended exam time — system flags scheduling conflicts and suggests adjusted courseload options automatically
  • 🛠
    Medical: concussion protocol Recent medical incident triggers coordinated outreach — professor notification, smaller section transfer, and TA support allocation
  • 📋
    Multi-advisor approvals Double major + minor requires sign-offs from 3 advisors and a professor exception — one shared view replaces running across campus for drop-ins
👥
Advisor-in-the-Loop
Supports conversations and decisions — never automates them. Advisors interpret, the system surfaces.
Strengths-Based
Surfaces interests, near-completions, and opportunities alongside risk — not just deficit tracking.
🤝
Equitable Access
Consistent early signals for all students reduce information asymmetry across caseloads.
🔒
Privacy-First
Operates within existing FERPA frameworks. Student data stays within institutional boundaries.