We connect the world of science with business
We implement solutions using, among others, Computer Vision, NLP, Audio Processing and Deep Learning.
Solutions using Machine Learning methods
We are working on the analysis of unstructured data, i.e. images, films, text, sound and various natural signals (e.g. galvanic skin response). It is particularly important for us to extract information that is useful for business purposes from the sea of data.
Most of our solutions use Deep Learning algorithms to classify information.
Some of our projects
CalfCam. Promity Agro AI.
CalfCam is an innovative AI-powered video monitoring system designed to improve calving supervision and herd profitability. It provides early alerts, helping farmers act faster, reduce losses, and enhance animal welfare.
Features:
- 24/7 automated monitoring of cows using AI-analyzed video feeds
- Early calving alerts — up to 30 minutes before delivery
- Real-time notifications via SMS, app, or call, accessible on mobile, computer, or TV
- Supports faster response, better care planning, and timely veterinary intervention
- Reduces cow and calf culling caused by calving complications
- Helps lower treatment costs and optimize staff work management
- Increases herd productivity and breeding profitability
- High detection accuracy: 95% effectiveness
Milk Profiler. Promity Agro AI.
MilkProfiler is an AI-driven solution that transforms milk quality assessment at delivery, enabling faster, more precise decisions. By combining microscopic imaging with advanced analytics, it ensures consistent quality and early pathogen detection.
Features:
- Evaluates milk deliveries in real time to support supplier pricing and acceptance decisions
- Analyzes microscopic images for:
- Fat globule size, shape, and distribution
- Mechanical damage and freezing/spinning indicators
- Recommends optimal processing based on milk quality data
- Detects pathogens using neural networks, including: E. coli, S. aureus, S. uberis, Candida, Prototheca
- Provides objective, lab-quality results in under 5 minutes — above 100× faster than traditional testing
Aimplant (a project for Aimedicum)
The project aimed to develop an application based on artificial intelligence, including Machine Learning and Computer Vision, to support surgical operations using a skull bone database. The application is designed for the reconstruction and restoration of lost skeletal fragments resulting from tumors or injuries.
Aimplant application is software for the automatic design of disposable medical devices using AI (i.e. implant templates) for maxillofacial surgeons.
Use case 1: After tumor resection, the bone was reconstructed using autograft in the form of fibula fragments.
Use case 2: As a result of an accident and surgery to remove a fragment of the mandible, it is possible to replace it with an implant or a fibula.
Access Control and Personal Protective Equipment
The PPE Detection application by Promity analyzes video frames from your camera devices (i.e. CCTV / VMS systems) in order to capture and monitor the occurrence of unsafe acts (e.g. people without vests or people without hard hats).
The software can be extremely beneficial for:
- Manufacturing sites
- Constructions sites
Key Features:
- Detection of people
- Ability to select a “region of Interest” that will be analyzed
- Possibility to analyze only frames on which motion has been detected, which significantly reduce consumption of computing power and internet resources
- Increase or decrease the sensitivity of the detection — you can adjust it to your needs hence avoid a large number of false positive alerts.
Cross-selling and up-selling insurance products
Based on classification models or recommendation engines, it is possible to
assign individual customers to offered products
along with the probability of achieving sales success. This will allow for better targeting of customers included in campaigns, thereby reducing campaign costs and increasing the return on investment. Moreover, secondary verification of customer segmentation based on recommendation results will be possible.
rFace
RFace is an image-based face recognition tool. It can operate in single or multi-person recognition mode. To identify a face, rFace is preconfigured (model training) by providing photos or videos with images of the people concerned.
The tool is dedicated to property security systems and personal safety. Thanks to automatic and instant image analysis in real-time, a person can be immediately picked-out of a crowd. rFace integrated with existing systems will ensure faster response and increased safety.
CX-Metrics. AI-driven Customer Experience.
CX Metrics is a system that uses data warehouses and artificial intelligence algorithms to obtain, constantly monitor and support the analysis of the characteristics and behaviour of large populations of people. The system can analyze about 40 personal characteristics, such as age, gender, emotions and selected features related to the appearance and lifestyle.
Perfect for companies which, based on CCTV systems, monitor the number of visitors to the place of sale, commercial and service outlets, shopping centres and large-format stores, as well as public places with an increased risk of undesirable behaviour, i.e. stations, airports, stadiums. The system meets the highest standards of data security, confidentiality policy and GDPR.
Latest blog articles on AI and Data Science
Our APIs got VERIFIED!
We are happy to announce, that our APIs on the RapidAPI marketplace have
We have integrated with RapidAPI!
If you’re looking for a high efficient real-time facial analysis, we’ve got
Shapley value in the context of machine learning
Motivation A topic that is often overlooked or underestimated by engineers or researchers,