Skip to content

Analysis Engine

Subscription Detection

Transform raw statements into subscription intelligence: recurrence, amounts, due dates, and alerts—with traceability, user control, and honesty about source data limits.

Detection is the foundation of ResiClic: without reliable initial mapping, any cancellation or negotiation starts from partial perceptions. We designed the engine as an enrichment pipeline rather than a black box: each subscription offered can be explained (rule, consideration, recurrence reason) and adjusted by you if ground reality differs.

∞
theoretical sources
banks, neobanks, manual exports
4
major cross signals
recurrence, amount, consideration, context
100 %
human control possible
merge, split, exclude by item
0 €
automatic bank reversal
we are not your bank

Processing Pipeline

Simplified diagram: in production, queues, validations, and safeguards complement these steps to handle import peaks.

Subscription detection pipeline schemaSourcesNormalizationDetectionAlertsActions

What we mean by 'detected subscription'

This is not a universal legal definition: it is an item classified as recurring commitment based on available signals—always revisable.

One payment can be a subscription, membership, installment payment, or labeling error. Our rules aim to isolate what behaves like recurring commitment (same amount at stable interval, known consideration, label consistent with a digital service or telecom). When ambiguity remains strong, we favor 'to validate' status over categorical assertion—precision beats the illusion of completeness.

  • Stable monthly payments from an identifiable provider (streaming, cloud, online press).
  • Trials or first months at reduced rate followed by full-price transition when the reason appears on the label or history.
  • Annual or quarterly plans with predictable due date.
  • Recurring 'add-on' options tied to a main service (additional storage, extra seats).
  • Telecom or box subscriptions whose label sometimes aggregates multiple components—detected as a grouped line with context note.
  • Embedded 'insurance' or tacitly renewed contracts when renewal signals are visible in the stream.
  • Classic B2C services hidden behind generic labels ('CARD PAYMENT INTERNET') with few clues—marked as 'low confidence'.
  • Edge cases: recurring donations, internal depreciation, leases—flagged rather than imposed as leisure subscriptions.

Six product levers to ensure mapping quality

Coverage & normalization

Consolidation by consideration, masked IBAN or aggregation, noisy label harmonization. We reduce 'Netflix' vs 'NFLX' duplicates when signal is sufficient.

Immediate budget reading

Amount per period, monthly equivalent, trend over 3 to 12 months, household subscription total contribution.

Contextualized alerts

Trial end, double charge, renewal increase, imminent annual switch—each alert cites the data triggering it.

Proof preservation

History snapshot useful for a complaint or cancellation: what was seen, when, at what confidence level.

Multi-account scenarios

Personal + joint + business account: filter by source without mixing legal frameworks or internal expense notes.

Resilience to partial imports

Error recovery, incremental import, reminder if a date range is missing—avoid temporal blind spots.

« Good detection is not what fills the table fastest: it's what warns you when it's uncertain. »

ResiClic Product Principles — Documentation interne

False positives, false negatives: how we handle them

Transparency on what algorithms cannot guarantee.

A false positive makes you think a service is still active when it is not; a false negative leaves a real subscription off the radar. Both cost time or money. We document confidence reasons (high, medium, low) and offer suggested groupings you can reject in one click. For merchants billing under various legal entities, detection can fragment the same offer: your manual merge teaches the engine for the coming months.

  1. Ingest

    Secure connection or file: we log the date range covered and parser version used.

  2. Classify

    Application of business rules + heuristic scores + merchant registry matching when available.

  3. Consolidate

    Cautious automatic merge; otherwise suggestion with summary 'why we thought it was the same'.

  4. Present

    List view, due date calendar view, household view—same data, three reading angles.

  5. Adjust

    Your corrections feed the local confidence profile (not shared with other users without legal basis).

  6. Re-evaluate

    At each new import, recalculation of alerts and projections without losing prior version history.

Typical sources and tradeoffs (indicative)

SourceAdvantagesUsual Limitations
PSD2 bank aggregationLess manual entry, fresh feedsDepends on connector and bank; some merchants missing
CSV / PDF exportFull user control, no API dependencyVariable quality; OCR sometimes needed; export delays
Multi-accountConsolidated household viewDouble-counting risk without merge discipline
Short historyQuick detection on existingAnnual or seasonal patterns less visible
Screen and cybersecurity concept

Sensitive data: scope and purposes

We limit storage to the strict minimum needed for detection, alerts, and preparing your actions. Third-party login identifiers are handled via token and rotation mechanisms compliant with industry standards. You can anytime revoke a connection or request export then deletion of your file subject to legal obligations (billing, ongoing disputes). Technical logs allow auditing who accessed what on the support team side—with internal traceability.

  • Systematic encryption in transit
  • Subcontractors listed in the privacy policy
  • No resale of transaction data to ad brokers

Audit mode

For accountants or family mediators: structured export of classification reasons and triggered alerts.

Service not found?

Submit a raw label to enrich our registry—prioritized manual handling.

Tax calendar

Annotation of potentially deductible items based on your situation—guidance, not tax advice.

Multi-currency

Indicative conversion with exchange rate date—useful for border workers and expats.

Usually within minutes for standard aggregators. Very large histories or complex PDF formats may need queue handling or user validation.

In extension, detection directly feeds the 'Manage & Optimize' modules (budget, tags, simulations) and 'Cancellation' journeys (proofs, timeline). We recommend complementing the digital view with a manual semi-annual pass: contracts under handwritten signature or payments on secondary accounts often escape automatic syncing.

From statement to decision

Discover how to budget, tag, and simulate cuts once your map is established.

Manage & Optimize

Back to home