AutosTimeline BR — Brazilian Case File Timeline & Pending Actions Builder

Send a set of case documents. Get back a correlated chronology with typed dates, document IDs, source pages — and a list of pending actions with no later record of compliance.

AutosTimeline BR reads multiple documents of the same case file (decisões, despachos, sentenças, certidões, petições, manifestações, atas, laudos) and builds:

• 📅 Chronological timeline of procedural acts, ordered by date with the origin of each date made explicit (filing, document, signature, act);
• 🆔 Document identification: type, court document ID (e.g. PJe/e-SAJ ID 2184355511), CNJ case number, and originating actor (juízo, Ministério Público, defesa, serventia);
• ⏳ Pending actions: orders and determinations (cite-se, intime-se, manifeste-se, apresente, cumpra-se) without a later record of compliance, with deadline, assigned party, and source;
• 👯 Duplicate documents (identical content or same document ID across files);
• ⚠️ Date inconsistencies (e.g. document dated after its own filing date).

⚖️ The product organizes the case file. It does not decide the case. It never concludes on prescrição, nulidade, autoria, materialidade, or mérito — every item links back to the exact excerpt, file, and page it came from.

Best for

• Law firms handling voluminous case files who need a fast, sourced chronology;
• Legal departments, insurers, audit firms, and investigative consultancies;
• Legaltechs and developers building case-management or legal-AI systems who need a structured timeline + pending-actions JSON as input.

Why not a generic document AI?

Generic assistants summarize one document at a time. AutosTimeline BR correlates documents: the certidão that fulfills a citação ordered two files earlier is what turns an "order" into a "possibly fulfilled" action. And dates are never merged into a single unexplained "event date" — the field used for sorting always declares where it came from (sortDateOrigin).

Input

Field	Type	Description
urls	string[]	Public URLs of the case documents (PDF, HTML, plain text)
documents	object[]	Pasted documents as {label, text} — for non-public files
caseReference	string	Optional case number; auto-detected (CNJ) when omitted
maxPagesPerDocument	integer	Page cap per PDF (default 100)
maxEventsPerDocument	integer	Event cap per document (default 100)

At least one of urls / documents is required.

Output

• Dataset — one row per timeline event (recordType=EVENT) or pending action (recordType=PENDING_ACTION):
  • EVENT: sortDate + sortDateOrigin (FILING_DATE / DOCUMENT_DATE / SIGNATURE_DATE / EVENT_DATE), event (normalized act, e.g. "receber a denúncia"), eventCategory (DECISION/ORDER/MANIFESTATION/HEARING/SERVICE/FILING), actor, documentId, documentType, all typed dates (documentDate, signatureDate, filingDate, eventDate, mentionedDates), attached deadline, sourcePage, sourceExcerpt, confidence, requiresHumanReview;
  • PENDING_ACTION: action, assignedTo, status (PENDING / POSSIBLY_FULFILLED + fulfillmentHint), originDocumentId, originDocumentDate, deadline, sourcePage, sourceExcerpt;
• SUMMARY.md — human-readable chronological report (open this first);
• timeline.csv / pending_actions.csv — spreadsheet-ready;
• OUTPUT — run counters: documents, pages, events by category, pending actions, duplicates, inconsistencies, errors.

Example

{
  "recordType": "EVENT",
  "caseReference": "0001234-56.2026.8.26.0100",
  "sortDate": "2026-05-04",
  "sortDateOrigin": "FILING_DATE",
  "event": "citar",
  "eventCategory": "ORDER",
  "actor": "juízo",
  "documentId": "2184355511",
  "documentType": "decisao",
  "signatureDate": "2026-05-04",
  "filingDate": "2026-05-04",
  "deadlineValue": 10,
  "deadlineUnit": "dias",
  "sourceExcerpt": "Cite-se o réu para responder à acusação no prazo de 10 (dez) dias.",
  "confidence": 0.85
}

{
  "recordType": "PENDING_ACTION",
  "action": "Manifeste-se o Ministério Público no prazo de 5 (cinco) dias. [prazo: 5 dias]",
  "assignedTo": "Ministério Público",
  "status": "PENDING",
  "originDocumentId": "2184400000",
  "originDocumentDate": "2026-06-10",
  "sourceExcerpt": "Manifeste-se o Ministério Público no prazo de 5 (cinco) dias."
}

How it works

1. Document layer (shared core with PrazoBR — the single-document deadline extractor): PDF page splitting (PyMuPDF), HTML/text loading, SSRF protection, per-page provenance;
2. Deterministic extraction per document: typed dates, document ID, CNJ case number, document type, originating actor, procedural acts with attached deadlines — no external calls, fully reproducible;
3. Correlation layer across documents: chronological sort with explicit date origin, duplicate detection (content hash + document ID), pending-action derivation (orders without a later compliance record), and date-inconsistency flags.

Every heuristic score is capped at 0.85 — ambiguous items are flagged requiresHumanReview instead of being silently guessed.

Limitations (MVP)

• No OCR: scanned PDF pages are skipped and counted in ocrPagesSkipped;
• The POSSIBLY_FULFILLED status is a transparent heuristic (later event by the assigned party, or a later certidão for citation/intimation orders) — it always ships with requiresHumanReview=true;
• Court-internal workflow acts (e.g. "conclusos para decisão") may appear as pending actions; treat them as "awaiting the bench", not party obligations;
• It does not calculate deadline end dates, prescrição, or any legal conclusion.

Companion product

Use [PrazoBR — Legal Deadlines & Obligations Extractor] for deep single-document deadline analysis; use AutosTimeline BR to correlate many documents into one case timeline. Both share the same document engine and JSON contract, so they compose in one pipeline.

FAQ

Does it work with scanned PDFs? Not yet — pages without a text layer are skipped and reported. OCR is not supported in the current version.

Does it send my documents to an LLM? No. The MVP is 100% deterministic: no external API calls beyond fetching the URLs you provide. Documents are processed in memory and never persisted.

Can I trust the chronology blindly? No — and that is by design. Every event carries its exact excerpt, file, page, and a confidence score, and ambiguous items are flagged for human review.

🇧🇷 Suite de Dados Públicos BR / BR Public Data Suite

Este Actor faz parte de uma suite brasileira de dados públicos e jurídicos. Combine-os em pipelines:

• LicitacoesBR — radar de licitações PNCP (editais e vencedores): https://apify.com/joaosbp/licitacoes-pncp-br
• CNPJ Lookup BR — consulta CNPJ em lote com dados da Receita Federal: https://apify.com/joaosbp/cnpj-enrichment-lookup-br
• CnpjDeltaBR — monitor de mudanças cadastrais de CNPJs: https://apify.com/joaosbp/cnpj-delta-monitor-br
• ComplianceBR — screening de sanções CEIS/CNEP/CEPIM: https://apify.com/joaosbp/cnpj-sanctions-screening-br
• Contact Scraper BR — emails, WhatsApp + enriquecimento CNPJ: https://apify.com/joaosbp/website-contact-finder-br
• PrazoBR — extrator de prazos e obrigações jurídicas: https://apify.com/joaosbp/legal-deadlines-extractor-br
• AutosTimeline BR — cronologia de autos e pendências: https://apify.com/joaosbp/case-timeline-builder-br
• PublicaBR — monitor de publicações processuais (DataJud/CNJ): https://apify.com/joaosbp/publicacoes-processuais-br
• EditalBR — extrator de editais de concurso em PDF: https://apify.com/joaosbp/edital-extractor-br
• Pricing Monitor BR — monitor de páginas de preço SaaS: https://apify.com/joaosbp/competitor-pricing-page-monitor

Pipeline sugerido: LicitacoesBR encontra editais → CNPJ Lookup enriquece o vencedor → ComplianceBR verifica sanções → CnpjDeltaBR monitora mudanças cadastrais.