MinervaAI transforms the practice of financial crime risk management to real-time risk assessment, with rich, relevant data, and predictive risk intelligence in a responsive construct. Technical briefing by Damian Tran, Jennifer Arnold, and Victor Tray.
Overview of Briefing
Damian Tran opens up the demo with a background on what MinervaAI is focused on and where this demo will take us.
We jump into the main user dashboard, which is currently in production. Programmatically, a user can establish a compliance program at scale and using MinervaAI’s API, a user can establish ongoing monitoring.
Damian takes us through an example of searching for an individual. We go through the required information for this search. As Damian explains, there are a number of pre-set data categories for users to select from to optimize their search. Selecting all the categories, Damian submits the search.
After the search is submitted, we wait for a real-time search process to complete. This is all done within seconds and we are presented with a knowledge graft, which is a quite vast and large set of data.
We see a number of profiles, with specific data clustered into their appropriate sections. Every profile has the key data that was found during the real-time search process.
While the report is loading, Damian dives into what MinervaAI does to reduce false positives.
With the e-report loaded, Damon shows us the data that is pulled during the real-time search process. We see a snapshot of risk information, which includes criminal, legal, news, and more. We see a lot of different sources used to find this information. All sources are cited properly for the user to check.
In the event that there is conflicting data, MinervaAI applies a consensus algorithm to verify the sources to find the correct data.
We move into the next tab, which is labeled “Watchlists.” We see different sanctions, criminal activity, and political exposure.
Next is the media tag, which we see have different levels and classifications. Two things are done here: sentiment analysis and classifying into several different categories. Additional rules and logic can be imputed to filter out the data as well. Links are provided for everything to allow the investigator to dive deeper.
In the event a link goes dead, MinervaAI provides screenshots for user documentation.
Damian shows us the rest of the sections that provide deeper information for the investigator. These sections are social and contact, links and ownership, and more.
Damian wraps up by showing the PDF report that is provided based on the data that was collected. Here we see screenshots of applicable sources to verify the data collected.
Damian wraps up the demo. Check out MinervaAI for more!