Idexa converts documents in different formats and structures into organized, comparable, and integrable data. It combines OCR (Optical Character Recognition, a technology that converts text images—such as photos, scanned PDFs, or printed documents—into machine-readable text that can be edited and searched), ICR (Intelligent Character Recognition, an evolution of OCR capable of reading handwritten text), OMR (Optical Mark Recognition, a technology that reads and interprets human-made marks), signature detection and verification, document classification, field extraction, and business rule application.

The workflow can receive images and PDFs, identify the document type, extract content, validate consistency, flag exceptions, and deliver data to ERPs (Enterprise Resource Planning systems), logistics platforms, financial systems, APIs (Application Programming Interfaces), and databases.

• Structured and unstructured documents: invoices, bills, production orders, receipts, forms, checklists, handwritten documents, and attachments in different formats.
• Validation: verification of mandatory fields, validity, signatures, markings, and customer-specific rules before the process moves forward.
• HITL – Human in the Loop: a model or process in which a human actively participates in the operation, supervision, correction, or decision-making of an automated or AI-based system. Teams no longer need to manually enter and fully review every file; instead, they focus only on exceptions identified by AI.
• Integration and scalability: structured data can feed logistics, financial, energy, insurance, BPO (Business Process Outsourcing), and industrial operations systems.
• Documented results: 99.5% accuracy; processing of 70,000 pages per month; 85% reduction in waiting time; 99% accuracy; and processing up to 25 times faster.

Idexa connects document automation to operational outcomes: it reduces manual data entry, queues, rework, and verification time while creating a structured data trail for analysis and integration.

Pix Safety transforms already-installed cameras into an active safety inspection network. Computer vision models continuously monitor areas, shifts, and routines, detect deviations, and provide evidence so that Health, Safety, and Environment (HSE) teams can act before unsafe behavior develops into an accident. The architecture leverages existing CCTV (Closed-Circuit Television) infrastructure and can incorporate new cameras and IoT (Internet of Things) devices, reducing the need for inspections conducted exclusively in person.

• Operational detection: detection of PPE (Personal Protective Equipment) use and non-use, cell phone use, handrail use, falls, access to restricted areas, exceeding occupancy limits, overcrowding, carrying objects, and proximity to hazardous zones.
• Alerts and traceability: event logging with image, location, timestamp, and the violated rule; forwarding to the platform and, depending on the configuration, through messaging channels.
• Preventive management: dashboards consolidate recurrence, critical areas, time periods, and types of deviations to support audits, training, and process improvements.
• Privacy and integration: the purpose is to classify safety-related events, not to identify workers by name. The solution can integrate with BI (Business Intelligence) platforms, corporate systems, and operation-specific rules.
• Expanded applications: access control, vital-sign monitoring in confined spaces, and revenue protection. Pix Thermo, for example, was a historical product developed for the healthcare sector and used during the COVID-19 pandemic to screen for fever in 2 to 6 seconds, with companies such as Volkswagen and Whirlpool.

The commercial offering combines deployment on existing infrastructure, 24/7 monitoring, and proven use cases with customers such as Petrobras and CEMIG.