The place where Data works for your Solution

Complex Machine Learning solutions for differents areas

About

We are an RnD team of extrovert Machine Learning nerds

Experience

The average experience of our team is 3+years in ML/AL project

Projects

We like challenging projects that require not only hard, but also creative work.

Product

We don’t believe that Al will destroy humanity, but we know how to biuld ML that will leverage your product.

Our experience

We have already cut our teeth on this areas: (from the deck p.4-5)

Numeric Data

Predictive modeling, exploratory analysis, classification, clustering, visualization, dimensionality reduction

Computer Vision

Image classification Object detection

NLP

Sentiment analysis, Text classification, Neural machine translation, Text summarization, Named entity recognitio

Solutions

TYPE OF THE PROJECT
Image classification model

WHAT WE HAVE DONE
The project was aimed to build the model to detect images with specific object. The project started In January 2018 and finished in February 2018. We completed data labeling phase, data exploration phase, creating convolutional network model, training, validation and deploy.

ТЕCHNICAL STACK
Python3

STAGE OF THE PROJECT
Completed

TYPE OF THE PROJECT
Image classification model (classic ML)

WHAT WE HAVE DONE
The project was aimed to build the model to detect logotypes on the web page. Project started in July 2017 and finished in December 2017. We completed data retrieval phase, labeling phase, feature engineering, data exploration phase, creating ensemble of models, training, validation and deploy.

ТЕCHNICAL STACK
Python3

STAGE OF THE PROJECT
Completed

TYPE OF THE PROJECT
Fraud prediction model for social network

WHAT WE HAVE DONE
The project was aimed to build the model to detect fraudulent activity of users. The model uses a data of users profiles, users activity, mobile data collected in created datawarehouse. Project started in May 2018 and finished in January 2019. We completed data collection phase, data exploration phase, feature engineering, choosing model, training and deploy.

ТЕCHNICAL STACK
Python3

STAGE OF THE PROJECT
Completed

TYPE OF THE PROJECT
Advanced dialogue system for Digital Assistants

WHAT WE HAVE DONE
The project was aimed to build the Dialogue system that can communicate with the user in a natural way. Project started in March 2019. For now we have completed research stage and we work on the enhancement of the working model.

ТЕCHNICAL STACK
Python3

STAGE OF THE PROJECT
On the go

TYPE OF THE PROJECT
GAN for Digital Assistants gaze simulation

WHAT WE HAVE DONE
The project was aimed to simulate the human like gaze movements in the digital avatars (assistants). The system was trained using the corpus of the videos with real human gaze movements for a narrow field of the human interaction context (greeting, basic emotions, etc.). Project started in October 2020 and finished in April 2021.

ТЕCHNICAL STACK
Torch-GAN (PyTorch)

STAGE OF THE PROJECT
Complited

Get In Touch

If you have any questions for us - let us know, we are happy to get something besides the error messages

Our stack

Python3

Rabbit MQ

Elasticsearch

Postgress

Spark

Spark ML

Sklearn

Pandas

Keras

Tensorflow

Google cloud

AWS cloudRabbit MQ