Article

Meet CRiSP: The cognitive risk-sensing tool that spots troubling trends

Mitigate risks to your brand

State of digital media

The scale of digital media is accelerating the global communication of public events. Opinions that once remained personal or localized are now being distributed to a worldwide audience with the click of a button. This abundance of digital data provides tremendous value for proactively identifying risks to brands, products, and services, but the colloquial language and phenomenal scale often prevents organizations from generating accurate and actionable insights. As a result, companies often fail to see a return on their investment in social media monitoring, whether through human capital or licensing tools.

Risk detection in digital media

Current techniques are incapable of keeping up with the rate at which new information becomes available. Some of the main shortcomings include:

  • Undetected risks: Legacy methods can miss risks that are undetected by keywords or because of the sampling methods
  • Untimely insights: Threats are not detected in a timely manner, which reduces the effectiveness of any mitigation that may have helped
  • Insufficient processes: Too much time is spent conducting risk identification and assessment activities

What is CRiSP?

CRiSP is Deloitte’s digital media-sensing solution, which uses customized machine learning models to monitor live digital media streams to bring the most troubling trends about your business to your attention. With a wealth of relevant data at your team’s fingertips, you are better equipped to generate responses quickly and to mitigate potential risk or damage to your brand. CRiSP uses state-of-the-art natural language processing (NLP) to understand how a person thinks and talks about your organization’s risks as well as artificial intelligence to address both language and scale.

Features and benefits of CRiSP

  • Language understanding: Using modern NLP techniques rooted in AI, CRiSP identifies risk related to keywords and factors in colloquial nuances of digital media language.
  • Flexibility: Goes beyond traditional sentiment analyses to capture areas of risk specific to your organization.
  • Self-learning: It uses feedback from your team to improve risk detection accuracy over time.
  • Scalability: It quickly processes structured and unstructured data from multiple digital media sources.
  • Data coverage: It monitors 100 percent of incoming data to locate needles in the haystack.

How does CRiSP work?

Step 1: Risk definition
Our industry professionals work with your organization to identify risk categories and customize the solution.

Step 2: Model training
Our team trains the NLP model by mining and tagging high-risk content.

Step 3: Data streaming
The NLP model streams and aggregates data from digital media sources into a single data warehouse.

Step 4: Risk tagging
The model monitors the data stream to tag potential risks flagged by ingested content.

Step 5: Analytics and alerts
The model automatically generates alerts when chatter concerning risk categories is increasing.

Step 6: User feedback
Your team can reject inaccurate results to teach the model to tag more accurately.

Step 7: Actionable insights
Supervisors have access to a wealth of data, which they can use to generate recommendations for the best actions to take to mitigate specific risks.

Contact us to request a demo or to learn more about CRiSP.

Did you find this useful?
/*Eloqua tracking*/
/*LinkedIn Insights tag*/