黑料不打烊

Platform

A model of your institution, and everything built on it

黑料不打烊 builds an ontology of your institution, then deploys applications, agents, and workflows on top of it.

  1. 1
    Connect your systemsSIS, CRM, aid, ERP, HR, LMS, advancement, and the documents around them
  2. 2
    Model your core business objects and relationshipsStudents, courses, aid awards, invoices, and gifts, related the way the institution works
  3. 3
    Encode the business logicRich expressions and ML models on the objects that matter
  4. 4
    Define allowable transactions as governed actionsEvery change runs through an action, inside each person鈥檚 permissions
  5. 5
    Keep the informal knowledge as memoryPolicies, advisor notes, emails, and past decisions, tied to the objects they explain
  6. 6
    Deploy on topApplications, agents, and workflows for every domain

Connect your systems

Every system, document, and outside source, in one place

Your systems
Student informationCRMLearning managementFinancial aidERP and ledgerHR and payrollAdvancement CRMService desk
Outside data
NCES and IPEDSCensusNational Student ClearinghouseBLS labor and wagesIRS 990sPeer public contracts
Your documents and history
PoliciesAdvisor notesEmails and inboxesContractsTranscriptsEvery past cycle and term

Model your core business objects and relationships

One model of the institution, across every domain

Each object belongs to a domain and links to the others. Logic and actions sit on the objects.

Institution ontology across seven domains: applicants, students, programs, courses, sections, accounts, aid, funds, donors, offices, revenue, and cashapplies tobecomespursuesfollowscalls foroffered astakesteachesownsemploysfundscollected intodiscountsopenshandled byplaced ontriggersbillspays intopaysgives toEnrollmentApplicantstageenumentry_termtermlikely_to_enrollpercentAcademic operationsProgramdegreeenumseatsintegerStudent lifecycleDegree plantarget_termtermcatalog_yearinteger茠on_trackbooleanAcademic operationsCoursecreditsdecimallevelintegernext_term_demandintegerAdd sectionInstitutional financeDepartmentcost_centerstringbudgetmoneybudget_variancemoneyWorkforceOfficedivisionenumstaffed_ftedecimal茠capacity_gaphoursStudent lifecycleStudentclass_levelenumcredits_earneddecimalpersistence_riskpercent茠unmet_needmoneyAcademic operationsSectioncapacityintegerenrolledinteger茠fill_ratepercentAcademic operationsInstructorrankenumload_targetdecimal茠assigned_loaddecimalInstitutional financeRevenue streamforecastmoneybudgetmoneyrevenue_forecastmoneyStudent lifecycleService casetopicenumstatusenumroot_causecauseStudent financeHoldkindenumblocks_registrationbooleanRelease holdStudent financeAccountbalancemoneydue_ondate茠balance_duemoneyStudent financeAid awardamountmoneystatusenumretention_liftpercentInstitutional financeCash positionbalancemoneyas_ofdate茠days_cash_on_handdecimalMove cash to investmentsAdvancementDonoraffiliationenumcapacity_bandenumlikely_to_givepercentAdvancementFundmarket_valuemoneypayout_ratepercent茠eligible_nowintegerAward from fundYour systemsOutside dataPolicies and memoryStudent information systemAdmissions CRMLearning management systemAid and billingERP and general ledgerHR and payrollAdvancement CRMAdvising and service deskNational Student ClearinghouseGraduate projectionsNCES and IPEDSMarket and interest ratesIRS 990 filingsCatalog and degree rulesHold, aid, and refund policiesGift agreementsDebt covenants and investment policyAdvising notes and email

A slice of the full model: the labeled objects are the ones this work runs on most.

Encode the business logic

Expressions and models that read across domains

Defined once on the objects, so every application and agent uses the same definitions and the same forecasts.

An ML model on Studentpersistence_risk reads grades, balances, holds, service cases, degree progress, and LMS activityAcademic operationsSection.midterm_gradegradeStudent financeAccount.balance_duemoneyStudent financeHold.blocks_registrationbooleanStudent lifecycleService case.statusenumStudent lifecycleDegree plan.on_trackbooleanStudent lifecycleStudent.lms_activitytrendML modelpersistence_riskChance the student doesnot return next term.Refreshed nightly,backtested on past terms.Student lifecycleStudentpersistence_riskUsed byAdvisor queueStudent services agentRetention outreach

It reads from three domains at once. Advisors and agents see the same score and the reasons behind it.

An expression on Fundeligible_now counts the students a fund's restrictions match who still have unmet needAdvancementRestriction.majorslistAdvancementRestriction.min_needmoneyStudent lifecycleStudent.enrolledbooleanStudent financeStudent.unmet_needmoney茠Expressioneligible_nowcount(students聽聽where restriction听听听听尘补迟肠丑别蝉聽聽and enrolled聽聽and unmet_need > 0)AdvancementFund茠eligible_nowUsed byAward from fundOfficer briefingsStewardship reports

It walks from a fund to its restrictions to the students they match, and uses unmet need, itself an expression from Student Finance.

Define allowable transactions as governed actions

Allowable transactions, defined once and enforced everywhere

Actions define who can change what, when approval is needed and where the change is written back. Permissions apply to every object and property.

Action 路 release_hold
On
Student 路 Hold
Who can run it
Bursar, student services agent
Approval
Second approver over $500
Writes back to
Student accounts
Logged
Who, when, why, and the evidence
Permissions 路 Student
AdvisorBursarRegistrarAgent
term_gpaYesNoYesinherits
balanceNoYesNoinherits
advising_notesYesNoNoinherits
holdsYesYesYesinherits

An agent inherits the permissions of the person it works for, and never sees more.

Keep the informal knowledge as memory

The informal knowledge stays with your model

Policies, notes, emails, and past decisions are tied to the objects they explain. Agents cite them when they answer. Memory adds context. It does not override the model.

HoldPolicy

Registration hold policy

Which balances block registration, and the exceptions the bursar allows.

ApplicationEmail

Counselor thread

What the family asked about, and what the counselor promised to follow up on.

Aid awardDecision

Appeal precedent

A past appeal and the reason it was approved, so similar cases are handled the same way.

Deploy on top

Frontier agents, an apps platform, and workflows, all on the model

Everything that runs reads the same objects and changes them only through governed actions.

Frontier agents

Agents that do the work, inside the rules

Agents built on frontier models read the model, reason with your policies, and act only through governed actions, with the permissions of the person they work for.

  • Read any object in the model
  • Act through governed actions
  • Cite the policy behind every answer
Apps platform

An application for every team, built from the model

Each domain ships with an application. New ones are assembled from the same objects, logic, and actions, with permissions built in.

  • Queues and object views
  • Forecasts with their reasons
  • Actions with approvals
Workflows

Workflows that combine agents, rules, and people

Triggers fire on changes in the model. Agents take the routine steps, rules route the rest, and people approve what matters.

  • Triggers on the model
  • Agent and human steps
  • Every step writes back through a governed action
Transfer credit evaluationWorkflow 路 Enrollment
Recent runs
yesnoTriggerTranscript receivedon any applicationAgentRead the transcriptcourses, grades, creditsAgentMatch equivalenciesrules, past decisionsRuleOn file?ActionPost transfer creditto the student recordNotifyTell the studentcredits, next stepsPersonChair reviewswith the evidenceActionApprove equivalencybecomes a rule on file

Continuous improvement

Results feed back into the model

  1. 1SeeEvery system and document in one model of the institution
  2. 2ActRoutine work runs automatically, and decisions stay with your staff
  3. 3MeasureResults tracked against goals, holdouts, and prior years
  4. 4ImproveForecasts and plans updated as results come in

Talk to us

Start with one domain, then expand on the same model

We scope the first domain with your team and agree on the outcomes to measure.

Talk to us How we deploy

黑料不打烊 is a part of , the operational intelligence platform.

Operations software for colleges and universities, built on one model of the institution.
contact@collegevine.com
Platform
Try an agent
Domains
Resources
Company
漏 2026 黑料不打烊. All rights reserved.