Skip to content
The four pillars

Four surfaces, one honest hand-off.

Cubico does the prep, research and drafting that used to cap how many students a counsellor could help. Four surfaces carry a student from a short intake to a briefed, human-led meeting, and a real counsellor stays accountable at every step.

Pillar 01 · Matchmaker

A fit score that shows its working.

The matchmaker compares a student against real programmes, at universities worldwide, on the things that decide an application: grades against entry requirements, total cost against budget, and the post-study work window a visa allows. It ranks and tiers the result, and it explains itself in plain words. In your Cubico account this is the Universities page.

  • Fit scored on real signals: academics, entry requirements, cost, and visa / post-study work.
  • Every match tiered reach · match · safety, so your matches are honest, never a wall of names.
  • A plain-word reason for each score: the AI renders the words, deterministic Python owns the numbers.
  • Verified, dated data at programme-and-round grain, with counsellor overrides winning over scraped facts.
  • Local realities are scored too: what a degree actually costs a family in Pakistan, and how deadlines land against board exams.
See how we handle cost and funding
Illustration: an example Cubico match. One university programme scores 88 out of 100 for fit, broken down into academic fit, entry requirements, affordability, and visa or post-study work, with a plain-language note explaining the score and a reminder that your counsellor reviews every match before it is final. The figures are illustrative, not a real result.
The line we never cross

The AI picks the words. Python picks the numbers.

The moment an LLM is allowed to invent a score, rankings drift between runs and trust collapses. So every number in Cubico (each fit signal, each tier) is computed by deterministic Python on verified data. The language model only turns those settled facts into a sentence a student and counsellor can read and challenge. Same match, same score, every time.

8
fit signals computed per match
3
tiers: reach · match · safety
0
numbers the LLM is allowed to invent
Pillar 02 · AI Counsellor

A tireless junior who never ships without you.

The AI Counsellor drafts the routine scaffolding (a first shortlist, essay outlines, a working timeline) so the human counsellor starts from a considered draft instead of a blank page. It runs on rails toward defined objectives and never acts on its own. Everything it produces is a proposal for a person to approve. In your Cubico account it is called AI counsellor.

  • A first-draft shortlist, essay outlines, and a timeline, assembled in minutes rather than evenings.
  • Outlines, never finished essays: the structure is suggested, the words stay the student's own.
  • It works on rails toward set objectives and never acts alone: every draft is a proposal.
  • Everything lands in the human counsellor's queue to review, adjust, and ship.
Pillar 03 · Chatbot

Answers a student can trust, and the ones it won't give.

The chatbot is read-only and grounded: it answers from the real programme dataset and the student's own profile, and it shows what it drew on. Just as important is what it refuses. Visa and immigration questions are regulated advice; loan and financing questions carry real money risk. It declines both by design and points the student back to their counsellor. In your Cubico account it is the Ask Cubico button.

  • Read-only answers grounded in the real programme dataset and the student's own profile.
  • It shows what it drew on (programme data, your profile) instead of asserting from nowhere.
  • Hard-declines visa-legal and loan-financial questions by design, and says why.
  • Hands anything needing judgement back to the human counsellor.
Pillar 04 · Briefing Engine

The artefact the whole product builds toward.

Matchmaking, drafts and grounded answers all feed one thing: the briefing pack your counsellor reads before your meeting. It is where a counsellor stops prepping and starts counselling, and it is what lets a single counsellor serve many students without cutting any of them short.

  • A structured briefing pack built from programme data, the student profile, and counsellor overrides.
  • Snapshot, fit rationale, tiered shortlist, deadlines, and talking points, all in one place.
  • Section-addressable: a counsellor edits any part, and each edit is tracked against the original.
  • Turns hours of prep into minutes, so one counsellor can serve many students well.
Where the human stays

Every pillar hands off to a person. None of them decides.

The matchmaker proposes, the AI Counsellor drafts, the chatbot informs, and the briefing engine assembles. The judgement call (which student applies where, and what the essay finally says) belongs to the counsellor and the student. Cubico compresses the work; it never takes the decision.

Universities worldwide

See it on your own matches.

Spend about 15 minutes on the intake and watch the four pillars go to work, with a counsellor beside you the whole way.