FOR IMMEDIATE RELEASE
Sydney, Australia – August 1, 2025
Vloggi, the smart video platform transforming how businesses collect and structure visual data, has launched a major upgrade that brings human-in-the-loop data labeling directly into the upload process, making video content immediately usable for machine learning pipelines.
With this new feature, contributors tagging their own videos during submission now deliver fully structured, legally-licensed video training data—dramatically reducing the time, cost, and risk involved in preparing datasets for computer vision.
“Our mission is to shift how the world sees video,” said Justin Wastnage, CEO of Vloggi. “It’s no longer just a creative asset, it’s a rich, structured data source. This release brings us closer to that future.”
A Foundational Step Toward vloggi.ai
This product update is the first milestone on the company’s roadmap to vloggi.ai, a newly soft-launched platform that allows anyone to build and monetize their own verticalized computer vision solution. Designed as a modular, no-code AI vision builder, Vloggi.ai enables resellers, industry experts and solution providers to launch branded video AI platforms tailored to their sector.
And at the core of that vision? The humble mobile phone.
“95% of businesses aren’t using computer vision—not because they don’t want to, but because fixed camera rigs and controlled lighting environments are prohibitively expensive,” said Wastnage. “We’re changing that.”
Mobile-First, Human-Smart: Accuracy That Competes with Static Cameras
Vloggi’s strategy centres on mobile video capture paired with human-labeled metadata—a potent combination that boosts usable AI accuracy dramatically:
- Raw mobile video: ~60% baseline accuracy
- Human input at upload: raises it to ~75%
- Pick-your-own CV models via Vloggi.ai: ~90% accuracy
- Advanced mobile capture (depth/LiDAR): goal of 95% accuracy by end of year
This closing accuracy gap is already reshaping sectors like infrastructure inspection, community reporting, training data collection, and incident logging, without the need for expensive hardware installs.
Ongoing research in partnership with Swinburne University of Technology is pushing the boundaries further, using next-gen smartphones’ LiDAR and depth-sensing capabilities to extract 3D video structure from ordinary footage. With co-developed and fine-tuned models tailored to specific industry needs, Vloggi expects its partner-led AI platforms to surpass traditional CCTV workflows in flexibility and affordability.
New Features Now Live
- Structured contributor fields — capture consistent metadata like region, category, department, or ID at point-of-upload
- Linked datasets — enforce dropdown validation with preloaded lists for things like store codes or project names
- Privacy-preserving blind whitelists — validate sensitive inputs (e.g. student ID) without revealing the full dataset
- Auto-tagging — automatically apply labels in the video library, streamlining post-upload filtering and sorting
About Vloggi
Vloggi is a modular, AI-powered video platform that transforms user-generated video into structured, searchable, and scalable data. The company is pioneering the next wave of human-in-the-loop video analytics, enabling businesses of all sizes to build and train vision AI without needing hardware investment. Vloggi’s platform supports everything from one-off video campaigns to fully-fledged white-labelled video intelligence systems. Founded in Sydney, Vloggi serves customers across government, enterprise, and innovation sectors.
Brand Comms Bureau
