Scroll down to discover more
Scroll down to discover more
Back

Metamaze, A Faktion Spin-Off: Intelligent Document Processing SaaS Platform

Metamaze:

A Faktion Spin-Off: Intelligent Document Processing SaaS Platform

Faktion launched Metamaze, an AI-driven IDP SaaS platform automating document processing at scale. With enterprise-grade ML, active learning, and seamless integrations, Metamaze enabled rapid growth—culminating in its acquisition by global data automation leader Duco.

Intro

Metamaze is an enterprise-graded AI-driven Intelligent Document Processing (IDP) SaaS platform that enables companies to automatically process, extract, structure and validate any kind of information from any type of document.

at

Need

Key Features

  • Data and Model Management: Seamlessly track, compare, deploy, and optimize models from training to production, allowing for data-safe model duplication and fast, multilingual training through incremental methods.
  • Fast & Easy Labelling: Effortlessly manage data annotation guidelines, providing step-by-step instructions for data entry users. Collaborate with team members for efficient labelling processes.
  • Active Learning for Enhanced Data Quality in Annotation: Enhance annotations with Active Learning, employ advanced machine learning for accuracy, and configure customizable human-in-the-loop processes with real-time oversight.
  • Easy Integrations and Monitoring: Integrate seamlessly with existing IT using APIs, webhooks, and out-of-the-box compatibility, featuring real-time monitoring through dashboards, user roles, and a detailed audit trail.

The solution

Data& Model Management:

Track,compare, deploy, monitor, and optimize models from training to production. Copymodels without sharing data. Multilingual, supports incremental training forspeed.

Fast & Easy Labeling:

Simplify data annotation with guidelines,step-by-step instructions, and team

Annotation QA with Active Learning:

Prioritize impactful documents forannotation. Detect and address anomalies to maintain solution quality.

Human Intervention:

Set quality thresholds and manage human-in-the-loop processes. Feedback improves deep learning models.

Configurable Pipelines:

Tailor production pipelines to yourneeds, validate data with business rules, and monitor in real-time.

Seamless Integrations:

Connect easily via APIs, webhooks, RPA, and integrations like SharePoint, Blob storage, FTP, and email servers.

User Roles, Audit Trail & Monitoring:

Manage user roles and maintain a detailed audit trail of all platform activity—whether performed by humans or AI—while accessing real-time dashboards that track performance, costs, savings, STP rates, operator productivity, and data quality.

Data& Model Management:

Track,compare, deploy, monitor, and optimize models from training to production. Copymodels without sharing data. Multilingual, supports incremental training forspeed.

Annotation QA with Active Learning:

Prioritize impactful documents forannotation. Detect and address anomalies to maintain solution quality.

Seamless Integrations:

Connect easily via APIs, webhooks, RPA, and integrations like SharePoint, Blob storage, FTP, and email servers.

User Roles, Audit Trail & Monitoring:

Manage user roles and maintain a detailed audit trail of all platform activity—whether performed by humans or AI—while accessing real-time dashboards that track performance, costs, savings, STP rates, operator productivity, and data quality.

Duco Acquisition

Faktion proudly announces the success of its spinoff, Metamaze,Metamaze Mock-up as it reaches a pivotal milestone with its international acquisition by Duco, marking a positive step forward in the global data automation arena.

Starting at Faktion in 2017 as the first sales hire, I eventually became Head of Sales and set up a highly efficient team. This led me to become CEO at Metamaze, a journey that has been incredibly rewarding for both myself and Jos Polfliet, my co-founding CTO and former Head of AI at Faktion.

Niels Van Weereld - CEO Metamaze

No items found.
No items found.

Outcome

10-40% more accurate

Metamazeis built on data-centric AI. By implementing active learning in the form offeedback loops and deep learning engines, our models become more accurate overtime.  

4x faster onboarding

Our data-centric AI approach also results in fasterproduction-ready models by only using the data that is needed to train themodel. This eliminates unnecessary training time and human effort.

507 Different Document Types Supported

2.947 Different Entities Supported

50+ Different Use Cases Supported

Related Cases

Discover all our cases