The Milestone Developer Summit 2025, dedicated to innovators, developers, and technology partners in video technology, took place at the historic Carlsberg Museum in Copenhagen—a venue renowned for scientific breakthroughs, including the invention of the pH enzyme, a milestone in biochemical innovation. This legacy of discovery provided the perfect global stage for a new era of open platform development, responsible AI, and, not least, the Hackathon.
Sebastian Döllner, VP of Technology Partnerships & Open Platform, said:
“Technology is evolving faster than ever, and no single company has all the answers. The winners are those who can bring together the best technologies, the best minds, and the best ideas. The essence of Milestone’s open platform is to empower innovation. We’re building more than video management software—we’re building an ecosystem that allows partners and customers to innovate on top of it.”
With keynotes from e
g. NVIDIA, AWS, Dell and Intel, breakout workshops and the Hafnia Hackathon, participants from all over the globe were able to share insights and predictions on what the future might look like.
First Access to the Future
Before the summit, AI developers from 15 different countries had early access to Milestone’s upcoming specialized VLM. The VLM is developed using NVIDIA Cosmos-Reason and post-training it with Hafnia’s domain- specific data library to better understand city-specific visuals, languages, symbols, events, weather, lighting and more.
The Hackathon challenge: create innovative integrations with third-party applications and leverage the VLM via API to enhance smart city solutions.
Today, millions of video cameras capture valuable footage that could improve everyday life. Yet much of this potential remains untapped due to the time-consuming task of reviewing and analyzing countless hours of video to uncover meaningful insights.
In addition to the specialized VLM, Milestone is also launching a generative AI-powered plugin for XProtect Video Management Software The upcoming generative AI-powered plugin for XProtect is designed to improve traffic management across cities, ports, airports, and other urban areas. It delivers advanced video intelligence by automatically converting video footage into detailed written reports, summaries, and validated real-time alerts.
And the Winner Is…
Of the participating developers, six were shortlisted, and a panel of judges from Milestone and NVIDIA selected the top three. The winner: Thomas Kreutz, who used Hafnia’s VLM API to turn live city cameras into instant, privacy-aware answers to real-time questions, said:
“I wanted to build something that is relevant and useful in real life
What I liked the most about the hackathon was how easy it was to get started. The API and documentation made it simple to build a demo quickly, brainstorm more ideas, and try them out.”
Thomas Kreutz took home €5,000 and the NVIDIA Jetson AGX Thor Developer Kit. Roland Harwood, Community Lead – Hafnia, commented:
“We are very impressed with the innovative integrations from all finalists. But Thomas Kreutz and Ask The City really nailed it. The hackathon success is promising for the future use of our platform and data library to train computer vision models on compliantly sourced, curated, extensively annotated, and anonymized real-world data.”
FACTS: The six Hackathon finalists ranked from winner:
1
Ask The City / Thomas Kreutz Turns live city cameras into instant, privacy-aware answers to real-time questions. Pick a place on the map, ask in natural language, and get a response derived only from the last few seconds of video. Powered by Hafnia’s VLM API.
- 2. VLM for Emergency Response / Flynn Owen, Dan Towers & Tom Aldridge (Blaize)
Leverages edge AI devices to generate events and contextualize, triage, and create response plans for emergency situations in near real-time.
- 3. SmartMap / Gabriel Crosnier de Bellaistre (Citilog)
Uses Hafnia’s VLM to unify incident detection, traffic levels, and weather/visibility data into live map
overlays with operator-ready actions, enabling faster decisions and better visualization.
4
RevoFlow / Rawinder Singh A no-code workflow builder that lets security integrators create AI-powered video analytics through drag- and-drop design, powered by Milestone’s VLM API. Audience Award winner.
- 5. Clipnotes / Magnus Guldberg Pedersen
Converts short video clips into structured, interpretable insights in seconds using Hafnia’s Vision-Language
Model.
- 6. Hafnia Video Event Analyzer / Jojy Saju JosephS
A multimedia intelligence platform that uses a VLM API to analyze images and video, generate semantic metadata, and enable advanced search, summarization, and event detection.