Methodology

EntryNest combines three layers:

1. A structured rules database — visa programs encoded as machine-readable requirements (age caps, language minimums, point formulas, financial thresholds, processing standards, document lists).
2. A deterministic scoring engine — runs the user's profile through the rule set and produces eligibility, risk, cost, and timeline outputs.
3. An AI generation layer — drafts personalized documents (SOPs, cover letters, interview answers) using the rule set as the source of truth.

The deterministic layer always runs first. AI never invents fees or eligibility rules; it personalizes content within boundaries the rule set defines.

The eligibility score reflects how closely a user's profile aligns with the published rules of a specific visa program.

• Hard requirements (age caps, degree level, language minimums, salary thresholds, financial proof) act as gates — missing one caps the score and the result is labelled ineligible for that program.
• Points-based factors (CRS for Canada Express Entry, points test for Australia, UK Skilled Worker tradeable points) are computed using the published government formulas, then normalized to the 0–100 scale.
• Soft factors (work experience depth, ties to destination, sponsor or job-offer strength, prior travel history) adjust the score within bands.

Scores are bucketed into Strong (75+), Moderate (50–74), and Weak (< 50). A strong score means your profile cleanly matches the program's published criteria — it does not predict an individual officer's decision.

Risk levels combine program-level signals with profile-level signals:

• Published refusal rates and category-level officer discretion from the destination government.
• Profile flags such as prior refusals, short employment history, weak ties, or document-authenticity concerns.
• Country-pair signals (visa-issuance trends for the applicant's nationality and country of residence).

Risk is a surface-area indicator, not a prediction of approval. It tells you which factors are most likely to draw additional scrutiny so you can mitigate them — adding stronger ties evidence, clarifying employment gaps, or addressing prior refusals — before filing.

The comparison engine evaluates every program the user is eligible for across three lenses:

• Fastest — based on the destination government's published processing standards plus biometrics and consular appointment realities.
• Lowest total cost — government fees plus typical third-party costs (language tests, ECAs, medicals, biometrics, IHS where applicable).
• Highest approval confidence — combination of eligibility score and risk level.

Rankings are computed live for the user's specific citizenship, residence, and intended purpose. They are not influenced by advertising, affiliate, or referral relationships.

Checklists are assembled from the official document list of the destination program and adjusted for the applicant's specific situation: family members included, prior visa history, country-of-residence requirements (police certificates, medicals), and program-specific extras (proof of funds amounts, language-test minimums, sponsorship letters). Each checklist item links to where the requirement is documented in the official source.

Timelines are sourced from official government processing-time portals (IRCC service standards, USCIS case-processing tool, GOV.UK service standards, Home Affairs visa-processing dashboard, etc.) and updated each review cycle. We display the published range rather than a single point estimate because real processing varies by case complexity, biometrics scheduling, and visa-office workload.

Cost ranges sum the published government fees plus typical third-party costs (language tests, ECAs, medicals, biometrics, translations, IHS where applicable). Amounts are shown in the destination's local currency with a USD approximation. Cost ranges do not include legal representation, relocation, or post-arrival expenses unless explicitly noted.

For document tools (SOP, cover letter, interview prep, refusal analyzer), AI generates personalized text within constraints supplied by the rules engine. Specifically:

• Prompts are pinned and versioned. The same input always produces a structurally consistent draft.
• Numeric facts (fees, thresholds, processing times) are injected from the rules database, not generated by the model.
• Output is validated against the program's known requirements — a draft that contradicts the rule set fails closed rather than reaching the user.
• Every draft is labelled as a starting point that the applicant must review.

• We do not predict individual case outcomes or guarantee visa approval.
• We do not give legal advice. Complex cases (prior refusals, criminal history, investor programs, family separation) should be reviewed by a licensed practitioner.
• We do not store the content of uploaded documents beyond what is required to return the requested output.
• We do not represent applicants before any government or tribunal.

For the full scope, see the disclaimer.