HiveCV

HiveCV is a Polish AI startup focused on edge and cloud computer vision solutions. The company was founded by Łukasz Piłatowski and operates a small technical team including Monika Osica (Manager) and Jakub Kowalski (Senior Software & Hardware Engineer). Their work centers on real-time image processing and data analysis, typically deployed at the hardware level using cameras from partners like Luxonis.

The company's output spans several verticals. In safety and logistics, they've built JanitorCV (person detection in monitored areas), SafeSense (alerts for excavator operators when people enter hazard zones), and Hard Hat & Vest detection for construction sites. In healthcare, Assisto detects raised hands above patient beds and alerts medical staff. They've also developed LabelGuard for package label verification, Age & Gender recognition software, and GestureLink, which translates hand gestures into device commands. On the interactive side, they created Artist 2.0, an exhibition combining Luxonis cameras with real-time generative AI, and are working on AI-powered virtual characters.

Services and capabilities

HiveCV positions itself as delivering edge AI and cloud AI solutions. The core capability is real-time image processing—detecting objects, people, gestures, and attributes in video feeds, then triggering alerts or actions. Most projects appear to pair vision models with hardware (typically Luxonis OAK cameras) and firmware to create turnkey systems rather than pure software consulting.

Their projects suggest experience with object detection (hard hats, vests, people), pose estimation (hand detection), and classification tasks (age, gender, label correctness). They work with generative AI for interactive applications like Artist 2.0. The company also mentions open-source projects and SDKs, suggesting they release tools alongside commercial work.

Engagement appears to be project-based. There is no mention of dedicated teams, staff augmentation, or time-and-materials contracts. The focus seems to be on building bespoke vision systems for specific safety or operational problems.

Notable work

JanitorCV detects unauthorized people in monitored areas and triggers alarms—used in facilities management. SafeSense addresses excavator safety by alerting operators when workers enter blind spots, a real liability issue on construction sites. Assisto targets hospital workflows, detecting patient calls (raised hands) and notifying nurses without requiring wearables or buttons. LabelGuard checks packaging quality by confirming labels are correctly applied, suggesting work in logistics or manufacturing QA. Artist 2.0 demonstrates more experimental capability: an interactive exhibition that responds to visitors in real-time using generative models. The Hard Hat & Vest detection project indicates compliance and OSHA-type safety applications.