Which AI would interest your business?


Video analysis for consumer experience enrichment


The business lens you need to better understand the past, the present and predict the future


Generative neural language system, capable both for conversation  and user real time conversation based profiling


Analysis of the user’s journey throughout a given online or offline ecosystem


Domain agnostic anomaly detection

Lummetry VAPOR+ is our advanced scene and video analysis solution based on custom Deep Vision technologies. Lummetry Video Analysis Processing, confidentialization – obfuscation and entity Recognition is a solution that targets multiple areas, ranging from GDPR-oriented basic features up to advanced video analysis. The solution is a complex AI Deep Vision powered video analytics platform capable of being deployed in various scenarios.

Key features:

  • Person tracking system for real-time or off-line video analysis
  • GDPR-compliant person obfuscation system based on selective actor tracking
  • Advanced human actor analysis including pose, emotion and demographics
  • Embedded storytelling capabilities of the video feeds, including configurable alerts for various purposes
  • Retail-specialized functions such as real-time merchandising planogram conformity analysis
  • On-demand customized video analytics features
  • Works either as a stand-alone online system (deployed on-prem or in-cloud) or as an API back-end

Lummetry VAPOR+ helps companies gather data automatically from their video feeds and then analyzes these data using Deep Learning modules in accordance with our clients’ needs. Its application ranges from security industry (e.g. real time burglary alert) to retail (e.g. planogram conformity) to healthcare (e.g. patient lying on the ground in clinic or homecare)

Lummetry LENS+ is our state-of-the-art AI predictive analytics solution designed for the retail industry but applicable for any industry with large number of clients, products or services (e.g. banking vertical, distribution & logistics horizontal, a.s.o.). Lummetry Latent-insights Entity analysis Neural System uses state-of-the-art Deep Learning approaches in order to create advanced knowledge graphs that act as a radiography of the analyzed business.

Key features:
Business radiography

  • Find products complex relationships (such as complementarities, needs addressing, etc)
  • Create usable content such as product bundling promotions based on the dynamic evolving patterns
  • Pinpoint at potential product cannibalization scenarios based on actual evolving consumer needs
  • Identify customer needs changes and construct insights accordingly

Actionable insights based on advanced Predictive Analytics

  • Demand & Sales forecasting based on complex non-linear patterns
  • Predictive stock replenishment
  • Predictive impact of custom actions (e.g. marketing campaigns, planogram changes, merchandiser actions)

Lummetry LENS+ helps companies understand the “buying behavior” of their consumers & partners when promotional campaigns are in place, when product seasonality occurs, or when customer behavior changes based on external factors.

Lummetry ALLAN is our generative neural language system, capable both for conversation (including in Romanian language) and user real time conversation based profiling. Moreover, ALLAN can perform intelligent tasks outside of chatbot type conversations such as analyzing documents and perform smart classification (e.g. emails’ classification), generate task-oriented specific insights and construct the writer’s profile. 

Key features:

  • Works on multiple areas ranging from online retail assistance up to back office document analysis
  • Interacts with external system in order to provide useful insights regarding human author profile
  • 100% Deep Neural based assistant. No rules or decision trees
  • Trainable on various tasks and texts corpuses
  • Analyses, understands and replicates humans’ natural language
  • Interacts with humans, being able to follow the conversational thread
  • Create user profiles based on the actual intents that can be determined within the conversation thread
  • Can be deployed as a online personal assistant or a back-office worker
  • Works either as a stand-alone online system (deployed on-prem or in-cloud) or as an API back-end

Lummetry ODISSEI is our product that can help you determine near real-time propensities of a certain action for a particular human actor (e.g buy a specific product or click a specific promotion in a online system, determine the employee probability of leaving a company and the actual timing), providing you have a detailed track record of specific users’ events (e.g. website tracking API for visitor events or employees’ track record details of activities). Moreover, it can determine, along the users’ journey, specific moments and different types of actions in order to increase the probability of reaching the desired target.

Key features:

  • Behavior-based user analysis including advanced clustering and meta-user segmentation
  • In-depth analysis of action & event ecosystem including behavior-based action segmentation
  • User journey state notification system (such as active, reactive, leaving in a working environment or engagement, disengagement, slow-engagement in an online system)
  • Insights generation at user-level or meta-user level
  • Predictive modelling of each user profile
  • Predictive modelling of event & action relationships
  • Works either as a stand-alone online system (deployed on-prem or in-cloud) or as an API back-end

Lummetry SHIELD is our product for domain agnostic anomaly detection and prevention based on actionable insights. Our Self-learning Heuristical Intelligent Entity anomaLy Detector is designed to work on a multitude of data streams in areas such as information security, banking risk, fiscal and transactional frauds, data confidentiality, industrial production anomalies, IoT systems, real time electronical/mechanical systems.

Key features:

  • Based on our own researched deep hierarchical siamese neural graphs the system
  • Analyzes various structured data such as accounting, security, user-information data-streams Uses advanced methods to analyze any kind of unstructured data ranging from raw text data up to scanned bitmaps document images
  • Successfully tested in various scenarios where the analyzed data stream contains both structured information and unstructured data
  • Implementation and configuration can deploy Lummetry SHIELD to various scenarios ranging from SIEM data analysis to banking transaction analysis
  • Analysis of data streams and detection of anomaly-like behaviors in real time streams of data or on Big Data repositories
  • Works either as a stand-alone online system (deployed on-prem or in-cloud) or as an API back-end