
Piotr Buczkowski
Founder of Xentim. A director and cinematographer from Poznan, Poland, formerly an art director at an interactive agency. He has worked professionally with images for years: video production, computer graphics, Unreal Engine and virtual production.
For the last few years he has worked with AI models that understand video and has built visual datasets to train them. He runs Xentim demos on clients' own footage.
Who we are
We are a team from Poznan that has worked professionally with images for years, and for the last few with artificial intelligence models that understand video. We build and test Xentim on our own cameras before we show it to a client.
What Xentim does
- Video analytics on existing cameras. A local device turns recordings into a list of events you search with a plain sentence, sends live alerts and a daily report. See: what is AI video analytics.
- Digital twins of places. A scan of a site as a 3D model you can measure and rearrange: an office scan in 104 objects and a courtyard from one camera. Pilots: emergency drills and factory layout planning.
- Synthetic data. Footage of events you cannot film: a fall, a fight, a person on the track, fire and smoke. Dataset catalogue.
How we work
- We test on ourselves first. Xentim runs on our own cameras before it reaches a client.
- The numbers in our articles come from our own measurements or from the sources listed under each text.
- The demo runs on the client's footage, and before permanent operation there is a trial period with measurement on their site.
Facts
- Based in Poznan, Poland.
- Analysis runs locally, recordings never leave the site.
- Own vision model, GOKON, trained for outdoor events.
- Xentim is a member of the NVIDIA Inception programme.
- Dataset samples: huggingface.co/xentim.
- LinkedIn: linkedin.com/company/xentim.
- Contact: contact@xentim.com.
See it on your own footage
The demo takes up to an hour and runs on material from your cameras.
Book a demo