Recommendations

Every report ends with a ranked list of recommendations: concrete, plain-language actions the applicant can take before submitting to improve their odds. Recommendations are not generic advice — they reference the specific gaps detected in the documents you uploaded.

How they are produced

Each red flag, missing-field, or weak-signal event surfaced by the analysis pipeline maps to one or more recommendation templates. Templates are then filtered by category, destination, and applicant profile, scored by expected impact, and deduplicated.

Anatomy of a recommendation

Field	Type	Description
id	string	Stable identifier you can use to dismiss or track.
priority	enum	high · medium · low — drives UI ordering.
category	enum	documents · finances · intent · history — matches the signal group it improves.
summary	string	One-line action the applicant should take.
detail	string	Optional longer explanation.
impact	string	Human-readable estimate of expected score improvement.
applies_to	string[]	Country codes for which this recommendation is relevant.

Example output

json

{
  "recommendations": [
    {
      "id": "rec_increase_funds_proof",
      "priority": "high",
      "category": "finances",
      "summary": "Add a savings statement covering the last 3 months.",
      "detail": "Your current statement covers 28 days; Schengen posts typically expect 90.",
      "impact": "+8 expected score points",
      "applies_to": ["DE", "FR", "ES", "IT", "NL"]
    },
    {
      "id": "rec_employer_letter",
      "priority": "medium",
      "category": "intent",
      "summary": "Attach an employer letter confirming approved leave.",
      "impact": "+4 expected score points"
    },
    {
      "id": "rec_book_return_flight",
      "priority": "low",
      "category": "intent",
      "summary": "Add a refundable return ticket on the booked date."
    }
  ]
}

Personalisation

Profile-aware. A self-employed applicant gets different finance recommendations than a salaried one.
Destination-aware. US B-1/B-2 emphasises ties-to-home-country; Schengen Type C emphasises insurance and itinerary.
Time-aware.If submission is < 14 days away, slow-to-obtain recommendations (e.g., notarised translations) are de-prioritised in favour of fast fixes.

Prioritisation logic

Recommendations are ranked by expected_score_uplift × feasibility_within_deadline. High-impact, easy-to-fix items always surface first.

Signal categories

Every recommendation belongs to one signal category. The category tells the applicant which part of their profile the action improves and drives how items are grouped in the UI.

Category	What it covers	Typical recommendations
`documents`	Completeness, legibility, and validity of uploaded files.	Re-upload a clearer scan; add a missing insurance certificate.
`finances`	Proof of funds, balance stability, and coverage windows.	Add a 3-month statement; explain a recent large deposit.
`intent`	Evidence the applicant will return and the trip is genuine.	Attach approved-leave letter; book a refundable return ticket.
`history`	Prior travel, visa history, and overstay/ties signals.	Disclose prior Schengen entries; add property or family ties.

Worked example

Consider a salaried applicant for a German (DE) tourism visa whose bank statement covers only 28 days and who has not attached an employer letter. The pipeline emits two events — a finances.coverage_short weak signal and an intent.no_leave_proof gap — which map to the first two recommendations in the response above. Because submission is three weeks out, both are feasible, so they rank by uplift: the savings statement (+8) sorts above the employer letter (+4). The refundable-ticket item is low priority and surfaces last. Acting on all three lifts the report from medium toward low risk.

Recommendations are not a guarantee

The impact figure is an estimated score uplift based on similar profiles, not a promise of approval. HeyVisa is not an embassy and cannot guarantee a visa decision — recommendations simply help you submit the strongest possible application.