Radiology's post-imaging layer
Laudos.AI is a Brazilian healthtech that builds software for radiology's most critical moment: when the image turns into a clinical decision, a report and secure communication. We were built by physicians who live the daily reality of the PACS and the on-call shift, and we exist to take weight off the radiologist's shoulders — not to take the physician out of the equation.
Why we exist
Radiologists spend too much time typing instead of analyzing. Hospitals demand volume and ship dated systems. And the critical finding — the one that must reach the referring physician fast — still travels by phone call and message, with no traceability or legal protection.
Our focus isn't the image, already well served by decades of PACS and viewers. Our focus is everything that comes after it: structuring the report, standardizing the language, suggesting the correct classification and making sure the right finding reaches the right person, on the record. That's why we call ourselves the post-imaging layer.
The stance is clear: this is not AI to replace physicians. It's AI to eliminate the work that shouldn't take up their time. The medical act — interpretation, validation and signature — stays human, as required by Brazilian Federal Council of Medicine (CFM) Resolution 2.454/2026.
What we build
The platform brings together three integrated modules covering the post-imaging cycle, always in assistive mode:
Copilot
Dictation and typing with automatic structuring by modality, guideline application and impression assembly. The radiologist speaks in natural language; the tool organizes technique, findings and conclusion for review.
LaudAI Search
Natural-language search for findings, hypotheses and patterns within the practice's context. Designed to speed up writing, always as support for the physician's decision.
CRIT
Flagging, notification and communication of critical findings with an SLA and documentation for legal purposes. Turns the untraceable phone call into an auditable record (available on the Enterprise plan).
See how the platform connects to your infrastructure on the integrations page, and the radiology terminology in the glossary.
Our principles
The physician in control
Every AI output is a reviewable suggestion. The radiologist edits, validates and signs. We never claim the tool diagnoses or replaces the professional — it speeds up the structure, not the clinical decision.
Built by people who report
Product decisions go through physicians who live the routine of the PACS, the on-call shift and the volume. If a feature doesn't save the radiologist real time, it doesn't ship.
Compliance is not a checkbox
Privacy, traceability and professional accountability are architectural requirements, not marketing ornaments. Data residency in Brazil, an audit trail and readiness for CFM Resolution 2.454/2026.
Connect, don't replace the workflow
We come in after the image and return the report to the existing workflow via HL7, FHIR and DICOM-SR. The practice doesn't switch PACS/RIS to use Laudos.AI.
Who's behind it
A lean team, close to the routine of those who report. Headquartered on Av. Paulista, São Paulo, Brazil.
Natan Paraíso Ribeiro
Founder · CEO and CTO · Data Protection Officer (DPO)
Radiologist · CRM-SP 192770 · Radiology training at InRad/HC-FMUSP
Founded Laudos.AI and ran the project solo until 2025. Leads product, engineering and clinical governance. As Data Protection Officer (DPO), he is responsible for privacy decisions, data residency in Brazil and compliance with LGPD/ANPD.
Francisco Akira
Clinical Lead · since 2025
Cardiologist · Practice at InCor
Joined in 2025 to connect the report's output to the referring world: what the clinician needs to read in the impression, how the critical finding should circulate, and what actually reduces the noise between image and care.
Timeline
- 2024 — Founding. Natan Paraíso Ribeiro, a radiologist, decides to tackle radiology's blind spot: everything that happens after the image — reporting, communicating and documenting — still depended on typing, the phone and memory.
- 2025 — Clinical lead and Portuguese terminology. Francisco Akira (cardiologist) joins as clinical lead. Copilot gains dictation with Brazilian radiology terminology, structuring by modality and suggestions of standardized classifications (BI-RADS, Lung-RADS, TI-RADS, Fleischner) in assisted use.
- 2026 — CRIT and institutional governance. Critical-finding communication with an SLA, audit trail, SSO and RBAC for imaging practices, hospitals and teleradiology. Ready for CFM Resolution 2.454/2026 and compliant with LGPD/ANPD.
Compliance and accountability
We treat privacy and professional accountability as architectural requirements. We operate in compliance with the LGPD, with a designated Data Protection Officer (DPO) — Dr. Natan Paraíso Ribeiro (CRM-SP 192770) — and a dedicated channel for data subjects.
- Assistive use: the tool supports writing; the physician reviews, edits and signs (CFM Resolution 2.454/2026).
- LGPD/ANPD: data residency in Brazil, a defined legal basis and data-subject rights handled by the DPO.
- Professional registration: company and technical lead with active CRM/CREMESP registration.
- Clinical evidence: classifications and recommendations grounded in peer-reviewed literature.
- Traceability: an audit trail with a timestamp and the user responsible for each action.
Frequently asked questions
Does Laudos.AI replace the radiologist?
No, and that's a non-negotiable stance. Laudos.AI is an assistive tool: the AI proposes structure, terminology and a draft impression, but the physician reviews, edits and signs. The tool does not diagnose or issue an autonomous report. The product is built for CFM Resolution 2.454/2026, which keeps responsibility for the medical act with the professional.
Who is behind the company?
Laudos.AI was founded by Natan Paraíso Ribeiro, a radiologist (CEO/CTO and Data Protection Officer, CRM-SP 192770), who ran the project solo until 2025. In 2025, Francisco Akira (cardiologist) joined as clinical lead. The headquarters are on Av. Paulista, São Paulo, Brazil.
Where is the data stored?
We adopt data residency in Brazil and processing in line with the LGPD, with a designated Data Protection Officer (DPO). For technical details on encryption, access control and retention, see the Security page and the Privacy policy.
Is the tool clinically validated?
The suggested classifications and recommendations are grounded in peer-reviewed literature and recognized guidelines (ACR, the Brazilian College of Radiology and specialty societies). Even so, every suggestion is reviewable and the final validation belongs to the physician.
How does an institution get started?
We offer assisted onboarding, with PACS/RIS integration, CRIT configuration and governance within two weeks. Talk to sales for a demo and to design the deployment project.
See the post-imaging layer in practice
Assisted onboarding: PACS/RIS, CRIT and governance within 2 weeks, with the radiologist always in control.