The Technology Powering Contrend

Contrend has been built on the latest, cutting-edge artificial intelligence (AI) and machine learning (ML) software. But that’s not all that powers Contrend. Keep reading to learn more about the data and technology behind our advanced MarTech SaaS.

Advanced Analytics

Contrend deploys advanced analytics that manage and interrogate massive data sets, allowing genuine AI and ML-driven decision making. These are not used in any content marketing MarTech platform or tech:

• Contrend’s unique, proprietary Data Lake
• K-Means Clustering—Used to identify similarities and differences in the huge data sets which allows for accurate calculation and application of EI scoring.
• DBScan—Used in parallel to K-Means.
• Longitudinal and Latitudinal data—Layered matrices that evolve over time.
• Randomisation—To shake up and refresh our automated harvesting bots, so they don’t get stuck in a scraping routine, and potentially miss relevant content.
• Random Forest—To develop and constantly improve Contrend’s automated ML recommendations and predictions.
• Bagging and Boosting—Used in parallel to Random Forest, B&B helps develop and constantly improve Contrend’s automated ML recommendations and predictions.
• PyTorch 1.9—Assists with visual treatment of layering Contrend data, applying image recognition methods, and deploying PyTorch’s new NLP capabilities.
• Computer Vision and Image Recognition principles

Sophisticated Tech Stack and AWS Hosting

Contrend’s tech stack is highly flexible, allowing for rapid and efficient API integration with client systems and third party software platforms. The platform takes advantage of a number of recognized, established programs and technology including Python, Elastic-Search, PostgreSQL, and more. All have been tailored and optimized for Contrend, and integrated with our multiple proprietary algorithms (patents pending).

As a certified independent software provider (ISV) with Amazon Web Services (AWS), Contrend also has secure cloud 2-foot mirroring in Singapore and the U.S.

Automated NLP Classification

Contrend has advanced unassisted learning capabilities that allows it to produce reports with unprecedented, high accuracy and with limited human training and intervention.

The platform gives rapid, accurate classification of multiple content criteria classes, including:

• Topic
• Style
• Format
• Visual Layout
• Use of Imagery
• Market of Origin

Proprietary Automated Harvesting Bots

The harvest bots that Contrend deploys ensure the complete content landscape is harvested for analysis with their self-regulating assessment of harvest depth and width. They are not driven by keywords, so Contrend identifies genuine audience preferences, not just those publishers that are good at SEO. And these bots are not limited by the layout of content, which means they can capture any webpage layout.

Through the power of our proprietary auto-bots, Contrend can deliver a rapid and accurate collection of multi-format URLs, based on a Research Subject. All it needs are some simple user inputs:

• Subject of interest
• Geographical market—One or multiple target countries or regions
• Owned URL—Optional, as Contrend can identify them, but when provided, it speeds up the onboarding process
• Competitors and URLs—Again, optional, but helps generate reports faster