Real-time media processing and AI video analysis at the edge

In high-traffic environments such as retail stores, transportation hubs, entertainment venues, and smart buildings, understanding real-time occupancy and pedestrian flow is critical for safety compliance, operational efficiency, and business intelligence. However, traditional methods, manual counting, turnstile-based systems, or basic sensor arrays tend to be costly, inaccurate, and unable to provide the spatial flow insights needed for informed decision-making.

A growing need exists for automated, camera-based solutions that deliver accurate occupancy counts, directional flow analysis, and zone-level breakdowns, all in real-time, without requiring personal identification, and in compliance with privacy regulations such as GDPR. With today’s advances in computer vision and AI, it would be possible to deploy intelligent video analytics systems that transform existing surveillance infrastructure into a powerful source of actionable data, processed entirely at the edge.

Web-based content consumption is growing rapidly, yet natural-sounding, browser-native text-to-speech remains largely unavailable for minority and co-official languages such as Catalan, Basque, Galician, and other European languages. News outlets, educational publishers, government portals, and cultural institutions need to make their digital content audible and accessible to all users, but existing TTS engines either lack support for these languages or produce robotic output that fails to engage listeners.

The EU Accessibility Act (effective June 2025) requires digital products and services to be perceivable and operable by people with disabilities, while WCAG 2.1 AA mandates that non-text content provides equivalent text alternatives and that content can be presented in different ways (including auditory) without losing meaning. Regional mandates such as Catalonia’s Llei de Política Lingüística add a further layer by requiring public-facing digital services to operate in co-official languages. Together, these frameworks create a compliance gap that most current TTS solutions cannot fill.

A browser-native AI text-to-speech solution would close that gap by enabling organizations to meet accessibility and language obligations simultaneously, while having a relevant social impact and helping digital publishers expand their audience reach.

Modern call centers and Business Process Outsourcers (BPOs) increasingly rely on video for customer service. Agents often work from home or in dense hubs using thin clients and virtualized desktops (DaaS & VDI). In these environments, video calls could expose distracting backgrounds and reduce perceived trust. A solution would require applying professional branding to all company agents’ screen backgrounds and allowing central management by the IT team without increasing server-side costs. However, integrating these solutions into DaaS & VDI environments poses a significant technical challenge due to the complexities of real-time video redirection and resource constraints.

In highly regulated sectors such as healthcare, finance, and government, the shift to remote work has introduced significant security liabilities. A remote worker’s webcam background could accidentally broadcast sensitive information, such as patient data on a whiteboard, confidential financial charts, or unauthorized personnel in the frame. Traditional server-side background removal solutions require raw, unedited physical video to traverse the network before AI processing, creating risks of accidental data breaches and regulatory non-compliance. While endpoint-based solutions can sanitize video before it enters the network, they face two primary hurdles: first, standard DaaS and VDI protocols do not natively recognize these processed streams as valid local camera devices; second, the intensive AI processing required often exceeds the computational capacity of thin clients and consumer hardware.