O
Oystr

Oystr, a Starian company founded in Curitiba in 2014, automates court work for Brazilian law firms and corporate legal departments. It has built more than 1,000 robots and run more than 500 million executions, capturing court updates, routing notices, monitoring hearing agendas, filing on court portals and moving data into case management systems around the clock. Oystr360 organizes notices by case, priority or type and can deliver them to Projuris ADV.

AI sits at specific steps. The Intimações app reads a notice and lays out its deadlines, decisions and observations, and the Protocolos app checks a filing's details against the petition and flags mismatches. Oystr Agent orchestrates AI agents and robots in parallel. It can detect a new suit, read the initial petition, classify it and register the case in the customer's ERP under configured rules, logging each action.

Mouzalas Azevedo automates more than 2,000 notices a month, and MDR Advocacia has used Oystr since 2022. In a joint deployment with Projuris and TechPro, Copel cut average case setup time by 90%. Every deployment is quoted, and an API runs single and batch jobs.

Capability grades

All 15 axes, graded from public sources on the date shown. Hover a grade to see what the letter means on that axis.

CC on AI CentralityArtificial intelligence is present but peripheral: a feature layer on a product whose value stands without it.

AI Centrality

How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.

Most of what Oystr sells is rule based court automation: more than 1,000 robots that collect notices, deliver updates, monitor hearing agendas, file on court portals and fill records in workflows the customer sets. AI enters at defined steps. The notices app adds an interpretation of each document's deadlines and decisions, and the filing app compares identifying details with the petition text to catch a discrepancy.

Oystr Agent, which Oystr calls the brain that commands the robots, uses AI to read a new filing and classify the event before rules decide the next step and robots carry it out. Routine capture, notification and filing status work runs without the models.

Source: Vendor Published
CC on Citation Accuracy and Hallucination DisclosureAccuracy is asserted without measurement, or grounding is claimed while output cites sources the reader cannot open and verify.

Citation Accuracy and Hallucination Disclosure

Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.

The notices app interprets a court notice that the user can open in the same workspace, and lays out its deadlines, decisions and observations beside it. The filing app compares the entered case number, parties, their roles and the file type with the petition and returns any discrepancy for correction. Both checks run against the case record and the source document. Oystr reports no measured error rate for notice interpretation, and the extracted statements carry no source citation of their own. No primary law corpus feeds the AI, which reads documents rather than forming legal conclusions.

Source: Vendor Published
BB on Autonomy and Oversight ModelA written commitment that the models work alongside a supervising lawyer, with real review surfaces, short of the full control structure: commonly the threshold at which the system stops or what happens after it is wrong.

Autonomy and Oversight Model

What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.

A user asks for an interpretation from the notice itself and checks it against the source document, and filing validation exposes differences between the petition and the entered data. Agent goes further. It decides the next step under the customer's rules and runs it in courts or systems, distributing tasks and updates with no handoff, and Oystr pitches it as work that needs no manual data entry. Every action is logged with who, what, when and the result.

In the Copel deployment a separate operations team handled exceptions and specialist review. No confidence rule stops an uncertain operation for a lawyer's review.

Source: Vendor Published
AA on Operational and Outcome EvidenceNamed firms or legal departments, dated, with figures for what changed and a method a reader can assess.

Operational and Outcome Evidence

Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.

Copel combined Oystr robots, Projuris and a TechPro operating team for case setup on a base of 210,047 files, 33,339 of them active. The combined workflow cut average case registration time by 90% and handled 10 times as many files in the same operating time. Mouzalas Azevedo Advocacia moved from manual handling to automating more than 2,000 notices a month, and its administrative director, Vanessa Porto, says staff now do only the work people must do.

MDR Advocacia has used Oystr since 2022 across notices, batch work and credential administration, and partner Weuder Martins says the product grew alongside the firm. Egger Mesquita, Mascarenhas, Goes and Inbursa also appear as customers.

Source: Vendor Published
CC on Privilege and Confidentiality PostureConfidentiality is asserted in general terms, or the commitment lives only in a sales conversation and cannot be read in advance.

Privilege and Confidentiality Posture

How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.

The privacy notice restricts personal data to authorized personnel, and organization users reach the workspace through their own accounts. A court notice, its interpretation and the outcome of each robotic step stay inside the customer's workflow. Oystr's data processing addendum lets it use processed data to improve its own and its partners' services and products and to prepare general statistics, preferably anonymized, with no exclusion for model training.

Court credentials sit in Presto, Oystr's separate encrypted vault with its own access controls. Attorney work product passing through an AI job receives no confidentiality treatment of its own.

Source: Vendor Published
BB on UPL and Professional Responsibility PostureA real position is published on advice versus tooling, short of full treatment: commonly a disclaimer without the supervision and competence dimension, or silence on jurisdiction limits.

UPL and Professional Responsibility Posture

Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.

Buyers are lawyers and legal departments that run court and filing workflows. The tools interpret court notices and compare filing details with the petition, and staff can open the underlying document, resolve mismatches and watch the workflow. Oystr's writing on AI in court work places responsibility for filed work on the lawyer and separates interpretation from configured court activity, and the Copel deployment routed exceptions to a human operations team.

Nothing addresses how supervision duties apply to a lawyer relying on Agent or on a generated notice interpretation, or how they vary by jurisdiction.

Source: Vendor Published
CC on AI Governance and Bias DisclosureResponsible AI principles are published without a mechanism, a testing regime, or anything a buyer could audit.

AI Governance and Bias Disclosure

Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.

Agent applies the customer's configured business rules after it interprets a document or event, and its audit trail records who acted, when, what happened and the result. The notice and filing apps keep outputs beside the source document for review. Oystr has written about the risk of errors when AI reads court documents, and its privacy notice lets a person ask for review of automated decisions that affect their interests.

Oystr names no owner for model assurance, no regular testing method and no result showing how performance varies across kinds of court notice.

Source: Vendor Published
BB on AI Safety and Data StewardshipSubstantive published policy covering most of the ground, short of the full set: commonly no named subprocessor list or no stated incident practice.

AI Safety and Data Stewardship

Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.

Oystr's data processing addendum, updated July 3, 2026, makes the customer controller and Oystr processor and requires notice within 72 hours of an incident that poses relevant risk to data subjects. Subprocessors are bound by written contract to Brazil's LGPD and listed by category, cloud hosting, restoration, IT consulting, incident response and AI based tools among them, with none named. After termination Oystr keeps data only for legal duties, audits or the defense of rights and then deletes it, and data in backups is made inaccessible until it can be removed.

The privacy notice keeps some digital identification data for six months and limits access to authorized staff. AI jobs carry no retention period of their own.

Source: Vendor Published
CC on AI Liability and RecourseLiability is addressed only through a standard limitation clause that disclaims the exposure the product creates.

AI Liability and Recourse

What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.

A liability cap sits in the data processing addendum, which forms part of every Oystr license and service contract. It limits Oystr to proven direct damage caused by Oystr or its subcontractors, up to the total contract value, and excludes lost profits and indirect damage except for gross negligence or willful misconduct. Liability for security incidents is limited to culpable or deliberate conduct by Oystr or its staff.

That cap is written for personal data processing. No warranty, indemnity or other remedy covers a customer that relies on a wrong notice interpretation or a mistaken automated court update.

Source: Operator Verified
AA on Practice Systems Integration DepthDocumented, verifiable integrations into the systems legal work already lives in, with the depth described: what syncs, in which direction, and what a firm must configure.

Practice Systems Integration Depth

How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.

Robots collect notice and proceeding information from Brazilian court portals, file on them, and put updates into the customer's case workflow. The Projuris ADV connection receives classified updates for the firm, and with Automações 360 an update can create a file, a task and a deadline and send it to the person responsible. The developer guides describe a single job API, a batch API and webhook connections. Agent reads a petition, extracts its data and writes it into the customer's ERP, keeping systems in sync in place of spreadsheet imports. Buyers can use those connectors alongside specific integrations for systems such as Legal One.

Source: Vendor Published
BB on Deployment Model and Data ResidencyDeployment model is stated clearly with partial residency detail, or residency is offered without the processing location being addressed, or the tenancy model is stated on its own with no residency detail published.

Deployment Model and Data Residency

Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.

The software runs as a hosted console whose robots exchange information with court portals and the customer's case systems around the clock. Oystr's privacy notice allows storage on servers in Brazil or another country where it uses hosted services, and processing abroad when cloud infrastructure requires it. The data processing addendum authorizes international transfers only to fulfill the main contract. A portfolio and its court portals can all be Brazilian while cloud processing follows a separate route.

Oystr names no hosting region, no choice of region and no location for the AI that reads notices or the output it stores.

Source: Vendor Published
DD on Security Certifications and Trust CenterNo independent security attestation located.

Security Certifications and Trust Center

Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.

Each automated action is logged, personal data access is limited to authorized people, and Agent is marketed with banking level security and no further detail. The data processing addendum gives the customer a remote audit on 30 days' notice, at its own cost and accompanied by Oystr staff. Presto stores law firms' and legal departments' court credentials in an encrypted vault with full traceability, a narrower control than the security of Agent and its AI interpretation.

Oystr holds no independent security attestation, such as ISO 27001 or a SOC 2 report, for the automation service or for the AI processing of notices and petitions.

Source: Operator Verified
CC on Model Supply Chain DisclosureThe vendor refers to advanced or proprietary models without identifying what sits underneath.

Model Supply Chain Disclosure

Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.

Agent places what Oystr calls specialized AI between court event inputs and the robots that act on the result, and coordinates several agents in parallel under central control. In the notices app the same kind of model interprets the source document inside the customer workflow. The architecture separates the part that interprets from the part that executes. Neither the model nor its provider is named, the data processing addendum lists AI based tools only as a subprocessor category, and no notice is promised when the model that reads court documents changes.

Source: Vendor Published
DD on Commercial TransparencyNo pricing information published at any level, including the unit of charge.

Commercial Transparency

Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.

Oystr quotes every deployment after a conversation with a specialist, scoped to the chosen modules, flows and integrations, and Agent flows are built to order for any operation that can be mapped to rules. Law firms and corporate departments buy different automation sets. No standard rate, unit of charge or subscription term is offered. The one public figure is a 2024 purchase by the Paraná water utility Sanepar, R$49,990 for specialized robotic services under a single contract, which sets no current rate for Oystr360 or Agent. Implementation and support charges sit inside each proposal.

Source: Operator Verified
BB on Firm and Practice CoverageSegment and practice coverage is described with substance, short of the boundaries: what is supported is clear, what is not is left open.

Firm and Practice Coverage

Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.

The tools are sold to Brazilian law firms and corporate legal teams that run litigation at scale, with Agent aimed at firms holding more than 5,000 active cases and at departments that need ERP integration and audit trails. Oystr counts more than 500 automated offices. Modules split the dispute operation into controllership, data entry, filing, case distribution, hearing agendas and notices, and firms use Oystr360 to collect and classify activity.

Customers include MDR Advocacia, Mouzalas Azevedo, Egger Mesquita, Mascarenhas, Goes and Inbursa, with work spanning civil collection, labor hearing calendars and utility case administration. The robots cover the courts and case systems they are built for, and no list of exclusions accompanies them.

Source: Vendor Published
Sources on file

7 public documents

The public pages on file for Oystr, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.

Pricing

No published figure

  • Oystr quotes every deployment and lists no standard price. The one public figure is Sanepar's 2024 purchase of specialized robotic services for R$49,990 under contract DL00866524. That single contract sets no current rate for Oystr360 or Agent.

Sanepar's 2024 purchasing register lists R$49,990 as the contracted amount for specialized robotic services from Oystr under contract DL00866524. The amount reflects one public purchase, not a list price for the current product packages.

Implementation: Custom workflow and integration work is scoped with the customer's specialist. The public purchase does not break out the setup charge.

Note: Oystr quotes the automation scope for each customer. A separate contract figure does not set a recurring starting rate for Oystr360 or Agent.

Legal Signals

What each signal means

A signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.

Confidentiality and Privilege

Client Data in Training

Can material a lawyer puts into this product be used to train a model?

Permitted, in the contract

The published agreement expressly reserves a right to train on customer content, with no opt out located. Any de identification, anonymization or aggregation qualifier is recorded in the summary.

The data processing addendum forms part of every Oystr license and service contract. Its clause 1.4(d) lets Oystr use processed data to improve its own and its partners' services and products and to prepare general statistics and studies, anonymized whenever possible. The clause contains no exclusion for model training, and the addendum mentions AI only as a category of subprocessor tool.

Source: Operator VerifiedMelhoria e aprimoramento dos serviços e produtos da Operadora e de seus parceirosAs of Nov 20, 2026Evidence

Prompt and Output Retention

How long does the product keep what a lawyer typed, and can that be set to zero?

Not addressed

No located public material states how long prompts and outputs are retained.

Some digital identification data is kept for six months, and after termination the data processing addendum keeps data only for legal duties, audits or the defense of rights before deletion. Generated interpretations stay in the notices app and Agent logs its operations. Prompts and model output have no period of their own.

Source: Operator VerifiedAs of Nov 20, 2026

Ethical Walls and Matter Segregation

Does retrieval respect the firm’s ethical walls, or can the model read across them?

Claimed, not documented

Segregation is asserted in public materials with no published detail on how it is enforced.

Personal data is open only to authorized people, and each customer runs its jobs inside its own workspace. Court credentials are held apart in the Presto vault with its own access rules. Nothing explains which matter restrictions the notice interpreter or Agent enforce when they read a document.

Source: Vendor Publishedsomente por profissionais devidamente autorizadosAs of Nov 20, 2026Evidence

Third Party Request and Subpoena Notice

If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?

Notice committed

Terms commit to notice where lawfully permitted. No transparency report located.

Clause 2.1(b) of the data processing addendum requires Oystr to inform the customer promptly of any communication or request from Brazil's data protection authority or another competent authority about the personal data it processes. The duty gives way only where a law or court order forbids notice. The privacy notice separately allows disclosure to judicial and administrative authorities when the law requires it.

Source: Vendor PublishedInformar prontamente à Controladora, salvo proibição legal ou judicialAs of Nov 20, 2026Evidence
Accuracy and Authority

Primary Law Corpus Provenance

Where does the law in this product come from, and does the vendor have the right to use it?

Jurisdictions only

Coverage is described by jurisdiction with no identification of the underlying corpus.

The robots capture official notices, filings and docket entries from Brazilian court portals for the buyer's own matters, and the AI interprets those documents. No separately licensed primary law collection stands behind the interpreter or Agent.

Source: Vendor PublishedAs of Nov 20, 2026Evidence

Good Law Verification

Does the product tell you when the authority it just cited has been overruled?

Not addressed

No located public material addresses whether authority is checked for subsequent history.

The robots report docket changes and route new filings through a workflow, and the AI reads the customer's own notices and petitions. Nothing checks how a later court treated a cited case or whether it is still good law.

Source: Operator VerifiedAs of Nov 20, 2026

Refusal and Uncertainty Behavior

What does the product do when the answer is not in the corpus?

Not addressed

No located public material addresses what the product does when it cannot ground an answer.

The filing app flags mismatches between the entered data and the petition. The notice interpreter has no described way to decline when it cannot find a deadline or read a notice with confidence, and Agent moves to the next rule based step without a confidence check.

Source: Operator VerifiedAs of Nov 20, 2026

Fabricated Citation Record

Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?

None located

No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.

The AI Hallucination Cases database kept by Damien Charlotin has no entry naming Oystr or Oystr Agent. The AI reads notices and petitions and extracts deadlines and case data rather than drafting filings with citations.

Source: Operator VerifiedAs of Nov 20, 2026Evidence
Professional Responsibility

Bar Guidance Alignment

Has the vendor engaged in public with the ethics opinions its buyers are bound by?

Generic reference

Public materials refer to professional responsibility in general terms without naming guidance.

The company's writing on AI in court work addresses lawyers' responsibility for what they file and sets it apart from automated court activity. It maps neither the notice interpreter nor Agent to OAB guidance, ABA Formal Opinion 512 or another named bar opinion.

Source: Vendor PublishedAs of Nov 20, 2026Evidence

Billing and Fee Posture

Does the vendor address what happens to the bill when the work takes an hour instead of six?

Savings claims only

Public materials claim time savings without addressing billing or disclosure, and the product sits inside a fee relationship between a lawyer and a client where those savings would change the bill.

Law firms cut the staff time spent capturing notices, entering data and checking hearing agendas, such as Mouzalas Azevedo's 2,000 automated notices a month. The workflow history records each action and its result. Nothing in the product shows how that work or its savings appear on a client's bill.

Source: Vendor PublishedAs of Nov 20, 2026Evidence

Outside Counsel Guideline Readiness

Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?

On request only

The material exists behind a sales conversation or an executed agreement.

The data processing addendum commits Oystr to provide the information and documents a customer needs to show compliance, and allows a remote audit on 30 days' notice. Subprocessors appear only as categories, cloud hosting, restoration, incident response and AI based tools among them. No model provider is named and no security certificate covers the service.

Source: Operator VerifiedAs of Nov 20, 2026Evidence

Court Disclosure Support

If a judge’s standing order requires an AI disclosure, can the product produce one?

Partial record

Some elements of the record are available, short of a document level export.

Agent's audit trail records who acted, when, what operation ran and its result, and the notices app shows the original notice beside its interpretation. Filing validation keeps the mismatches it caught. No single export brings the model, its sources and the human check together for disclosure to a court.

Source: Vendor PublishedAs of Nov 20, 2026Evidence
Contact

Correct a record, or ask how something was graded

Every grade and every signal on this index is drawn from public sources and dated. If a record is wrong, out of date, or missing an artifact the index did not locate, send the source and it will be reviewed and the record redated. Vendors are welcome to submit documentation. Nothing on this index is for sale, including a listing, a placement, or a grade.

AI Legal Index

The AI Legal Index is an independent index that tracks changes to AI vendors in legal. It holds 617 vendors across 9 categories, each graded on the same 15 capability axes and recorded against 12 legal signals, from public sources. No vendor pays for inclusion, placement, or rating.

Index Status
Last index update
November 20, 2026
The AI Legal Index is an editorial reference. It is not a regulatory body, not a law firm, and nothing published here is legal advice or a recommendation to retain or avoid a vendor. Records are verified against published sources, bar guidance and public court records. Where a record reads not addressed, the material was not located in public sources on the date shown. See the Methodology page for evaluation standards and limitations.
© 2026 AI Legal Index
3801 N Capital of Texas Hwy, Ste E240 · Austin, TX 78746