Spectrum Labs for Engineering

Behavior detection at the scale and speed of the internet

Groundbreaking technology to detect a range of human behaviors on platforms that host user-generated content.

Schedule a Call

AI for Understanding and Driving Human Behavior

Spectrum Labs’ AI leverages transformer neural networks alongside robust data workflows to detect both healthy and toxic behaviors in online communities.

Our data vault is the world’s largest data set specifically built for training AI to detect both positive and negative behaviors in online platforms – and we use it to help our customers keep their communities safe and strengthen the user experience on their platforms.

Understand and drive behavior in your community

Transformer neural networks

Transformer models are well-known through the large language model (LLM) applications that are currently taking the world by storm.

Spectrum Labs has leveraged language models for years. XLM-RoBERTa is a transformer model used at Spectrum Labs that is trained on large human-labeled data sets compiled specifically for capturing a full range of human behavior in user-generated content (UGC).

341_SL_ABS_DataVaultChart_LV1_031323_OPT2

Years of AI R&D ready-to-go

Advanced behavior systems

Spectrum Labs leverages transformer neural networks to drive determination of toxic behaviors at scale. We’ve been doing this over several years for the world’s largest UGC platforms.

We offer more than just “one size fits all” models. Our Advanced Behavior Systems (ABS) are a fully configurable workflow for parsing human behavior at scale using a combination of XLM-RoBERTa models, wide-ranging data sources, and specialized data operations with active learning cycles.

Learn More About Advanced Behavior Systems

Action content before it gets posted

Scalable, real-time API

When it comes to capturing the type of content and behavior that is posted on UGC platforms, time is of the essence.

When dealing with illegal content, action needs to be decided instantly – not within days, minutes or even seconds.

Spectrum Labs’ API returns a full determination for action within 20 milliseconds. Our API processes tens of thousands of messages per second (billions a day), and has scaled coverage across the world’s largest UGC platforms.

Apply OOTB real-time decisions to your platform

Real-time decisions and event-based actions

Spectrum Labs’ API not only provides behavior determination in real-time, it also comes with an action framework.

The API response includes, alongside the behavior identified, tags with possible actions to perform as well as previous behaviors captured from the user. This allows for real-time (under 20 milliseconds) automated decisions on content, which offers great value to our customers in terms of reduced engineering costs and time.

Additionally, the API provides event-based and rule-based webhooks that allow for highly nuanced action automation on content and users. These highly flexible webhooks enable our customers to achieve truly complex automation workflows at a very large scale.

EASY IMPLEMENTATION AND ACTIONABILITY

Easy API, decisioning, and webhooks

Implementing Spectrum Labs' solutions is easy through our well-documented API and webhooks which require only minimal engineering resources to get up and running.

Our API comes with a real-time decisions framework where you can configure complex business rules around the actions taken when a prompt or output is in violation of your policy. The API response will return a determination of the detected behavior and the action to be taken on it within 20 milliseconds.

Additionally, our event-based action framework allows you to set complex rules and fire off a webhook once those rules are met, allowing for complex workflows.

Spectrum Labs prides itself on its first-class customer support. All of our clients are provided with a dedicated solutions consultant who works closely with your organization from day one to help oversee the entire implementation phase.

Learn More