AI | Machine Learning | MLOps

We create and implement solutions powered by Artificial Intelligence, enabling sustainable competitive advantages in the Digital Era

Artificial Intelligence | Machine Learning | MLOps |

We create and implement solutions powered by Artificial Intelligence, enabling sustainable competitive advantages in the Digital Era

Automation of ML-models lifecycle

Neoflex MLOps Center solution provides easy-to-use and efficient infrastructure to automate the work of data science teams and enable quick and easy deployment of ML-models to production

DOWNLOAD PRESENTATION

CONTACT

Automation of ML-models lifecycle

Neoflex MLOps Center solution provides easy-to-use and efficient infrastructure to automate the work of data science teams and enable quick and easy deployment of ML-models to production

DOWNLOAD PRESENTATION

CONTACT

Projects

DEVELOPMENT OF PD, LGD MODELS AND MACRO ADJUSTMENT FOR BANKS

Customer

A number of banks in Europe, Africa and CIS countries

Task

To comply with the requirements of regional and European regulators in terms of the IFRS9 financial reporting standard, the banks should develop models for assessing the borrowers' credit risk: default probability models, loss given default models and macroeconomic adjustment models.

Result

A developed stack of models estimating observed default rates (ODR) using a vintage approach, a number of implemented algorithms for fitting parametric curves to ODR (including Weibull-Pareto, Cox PH, Gen Gamma, RF, etc.). LGD models, automated backtesting procedures, and macroeconomic adjustment models (regression models and SVMs) have been developed. A series of successful external audits of Big-4 companies (EY, PWC) has been conducted.

FORECASTING WAREHOUSE STOCK VOLUMES

Customer

One of the largest Russian logistics companies

Task

Implement a model for forecasting the warehouse stock volume in all departments of the logistics company to increase the accuracy of transport resources planning.

Results

Neoflex has developed a set of time series models using various approaches (STL + SARIMA, ARIMA+Fourier, TBATS, Prophet, LSTM). The SARIMA model has shown the best results; median weighted percentage error is ~28%. The accuracy of the developed planning algorithm has allowed the planning accuracy to be increased by more than 5% compared to the previously used algorithms.

SEARCHING FOR ANOMALIES IN NETWORK LOGS

Customer

Top-10 Russian Bank

Task

Implement a model enabling identification of hidden anomalies related to the informational security events, which arise due to network attacks.

Results

The model is implemented in the form of a containerized service that enables searching for anomalies both in batch and online modes. The model is based on an entropy approach (the Shannon entropy is used as a measure) that detects network traffic anomalies the standard SIEM tools may not respond to (for example, multi-vector or 'slow' DDoS attacks).

EMPLOYEE CHURN FORECASTING MODEL

Customer

A large Russian IT company

Task

To reduce the turnover of front office employees (developers, testers, system analysts), we need to identify a set of key factors affecting a significant increase in the probability of employee termination.

Results

A developed model using AutoML approaches (the best result was obtained using the Distributed Random Forest algorithm), AUC = 0.64 and recall of terminated employees is approximately 0.68. This model helps identify employees who are in the "risk zone" based on more than 50 factors which, in turn, enables vertical managers and HRs to specifically intensify actions to retain employees and take timely measures.

CONTACT US

Projects

DEVELOPMENT OF PD, LGD MODELS AND MACRO CORRECTIONS FOR BANKS

Customer

A number of banks in Europe, Africa and CIS countries

Task

To comply with the requirements of regional and European regulators in terms of the IFRS9 financial reporting standard, the banks should develop models for assessing the borrowers' credit risk: default probability models.

Result

A developed stack of models estimating observed default rates (ODR) using a vintage approach, a number of implemented algorithms for fitting parametric curves to ODR (including Weibull-Pareto, Cox PH, Gen Gamma, RF, etc.). LGD models, automated backtesting procedures, and macroeconomic adjustment models (regression models and SVMs) have been developed. A series of successful external audits of Big-4 companies (EY, PWC) has been conducted.

FORECASTING THE ACCEPTED CARGO VOLUME

Customer

One of the largest Russian transport companies

Task

Implement a model for forecasting the accepted cargo volume in all departments of the logistics company to increase the accuracy of transport resources planning.

Results

A developed set of time series models using various approaches (STL + SARIMA, ARIMA+Fourier, TBATS, Prophet, LSTM). The SARIMA model has shown the best results; median weighted percentage error is ~28%. The accuracy of the developed planning algorithm has allowed the planning accuracy to be increased by more than 5% compared to the previously used algorithms.

SEARCHING FOR ANOMALIES IN NETWORK LOGS

Customer

Top-10 Russian Bank

Task

Implement a model enabling identification of hidden anomalies related to the informational security events, which arise due to network attacks.

Results

The model is implemented in the form of a containerized service that enables searching for anomalies both in batch and online modes. The model is based on an entropy approach (the Shannon entropy is used as a measure) that detects network traffic anomalies the standard SIEM tools may not respond to (for example, multi-vector or 'slow' DDoS attacks).

EMPLOYEE DISMISSAL FORECASTING MODEL

Customer

A large Russian IT company

Task

To reduce the turnover of front office employees (developers, testers, system analysts), we need to identify a set of key factors affecting a significant increase in the probability of employee termination.

Results

A developed model using AutoML approaches (the best result was obtained using the Distributed Random Forest algorithm), AUC = 0.64 and recall of terminated employees is approximately 0.68. This model helps identify employees who are in the "risk zone" based on more than 50 factors which, in turn, enables vertical managers and HRs to specifically intensify actions to retain employees and take timely measures.

CONTACT US

Our solutions

ML-models for financial credit risk management

We develop and implement models for assessing credit risks in the financial sector (PD, LGD, early warning models, macroeconomic models for IFRS9)

Accurate forecasts and planning optimization

We develop services for forecasting multidimensional time series (sales volumes, warehouse loading, information systems load) using ML models

Detection of the hidden threats

We create services for detecting anomalies based on ML models

ML-driven HR management

We develop and implement models for the HR domain: forecasting employee termination and automating the search for employees based on ML-models

Customer retention, churn handling and cross-sell recommendations

We develop ML-models of customer churn, NBA/NBO models, AI-based recommendation systems

NLP for improved operational efficiency

We develop natural language processing (NLP) services for processing customers' requests, news information, automated parsing and verification of text documents

ML-models lifecycle management

We build an easy-to-use and efficient infrastructure to automate the work of data analysts and enable mass integration and operation of ML-models based on the Neoflex MLOps Center solution

Cloud-Native services based on ML models

We develop high-performance cloud services for ML model inference based on the infrastructure of leading providers: AWS, GCP, Yandex.Cloud, SberCloud

Our solutions

ML-models for financial credit risk management

We develop and implement models for assessing credit risks in the financial sector (PD, LGD, early warning models, macroeconomic models for IFRS9)

Accurate forecasts and planning optimization

We develop services for forecasting multidimensional time series (sales volumes, warehouse loading, information systems load) using ML models

Detection of the hidden threats

We create services for detecting anomalies based on ML models

ML-driven HR management

We develop and implement models for the HR domain: forecasting employee termination and automating the search for employees based on ML-models

Customer retention, churn handling and cross-sell recommendations

We develop ML-models of customer churn, NBA/NBO models, AI-based recommendation systems

NLP for improved operational efficiency

We develop natural language processing (NLP) services for processing customers' requests, news information, automated parsing and verification of text documents

ML-models lifecycle management

We build an easy-to-use and efficient infrastructure to automate the work of data analysts and enable mass integration and operation of ML-models based on the Neoflex MLOps Center solution

Cloud-Native services based on ML models

We develop high-performance cloud services for ML model inference based on the infrastructure of leading providers: AWS, GCP, Yandex.Cloud, SberCloud

Why Neoflex

1

A unique combination of engineering competencies (Big Data, CEP, MLOps, development) and vast expertise in ML modeling

2

We solve both problems related to the development of ML models and their adaptation to real conditions, integration into a production run, and subsequent monitoring of the quality of their operation

3

We use top-class software products for automating ML processes: MLflow, Kubeflow, DVC, etc.

4

Our Cloud-Ready solutions are based on Kubernetes or OpenShift, which makes them easily portable and scalable

5

Our Data Scientists and Data Engineers have deep expertise in developing ML/AI models and creating analytical solutions. Their experience, combined with the vast practical experience of DevOps engineers, can dramatically increase the speed of integration of ML/AI models.

6

All our projects are implemented by a highly qualified team of data engineers, ML engineers, domain experts, Data Science and MLOps experts

CONTACT US

Why Neoflex

1

A unique combination of engineering competencies (Big Data, CEP, MLOps, development) and vast expertise in ML modeling

2

We solve both problems related to the development of ML models and their adaptation to real conditions, integration into a production run, and subsequent monitoring of the quality of their operation

3

We use top-class software products for automating ML processes: MLflow, Kubeflow, DVC, etc.

4

Our Cloud-Ready solutions are based on Kubernetes or OpenShift, which makes them easily portable and scalable

5

Our Data Scientists and Data Engineers have deep expertise in developing ML/AI models and creating analytical solutions. Their experience, combined with the vast practical experience of DevOps engineers, can dramatically increase the speed of integration of ML/AI models.

6

All our projects are implemented by a highly qualified team of data engineers, ML engineers, domain experts, Data Science and MLOps experts

CONTACT US

Techstack

Our customers for AI/ML and MLOps solutions

About Neoflex

Neoflex creates IT platforms for a digital business transformation which helps our customers gain sustainable competitive advantages in the digital era.

We focus on custom software development and implementation of complex information systems using сutting-edge technologies and modern methodologies.

Our industry and technological expertise, enhanced by our own software accelerators, allow us to solve business tasks at any level of complexity. Among our customers are more than half of Russia’s Top-100 banks, 11 bank of the world’s TOP-100, as well as companies from 22 countries in Europe, Asia and Africa.

LEARN MORE

Tell us about your tasks

+7 (495) 984 25 13

О Neoflex

Expertise

AI | Machine Learning | MLOps

﻿Our solutions

﻿Our solutions

Our solutions

Our solutions