Back to Portal

Investment - August 2021

The Janvest General Partnership is pleased to announce the execution of Limited Partnership IV’s sixth investment in an Israeli founded emerging technology company. Below you will find information about the business and corresponding investment. We welcome your questions and feedback.

DEAL SUMMARY
Company Name
/ TRSTai
Website / Under Construction
Offices / Tel Aviv
Sectors / Machine Learning Operations (MLOps)
Founders / 2
Year Founded / 2021
Number of Board Positions Held by Janvest / One
Total Janvest LP IV Investment / $1,000,000
Total Size of Round / $2,000,000
Notable Co-Investors / Tensility Ventures (Chicago), Maccabee Ventures (NYC), SeedIL (Israel)
LP IV Ownership / ~14.0%

           

                   

                                                                                                                                                                                                                                                                                                                     View fullsize                                    

         

   

 TECHNOLOGY DESCRIPTION
TRST is bringing to market the first real-time auto-audit and remediation solution for Machine Learning (ML) models. These models are the brains behind an ML solution’s ability to effectively process vast amounts of data and find high value correlations with corresponding predictions. The challenge with rolling out production-grade Machine Learning is that so many of these models underperform due to flaws and deficiencies inherent in the data they process. As a result of this flawed data, ML models tend to degrade over time thus resulting in Bias (systematic prejudices), Drift (inaccuracy of predictions), Security Manipulation (false predictions through malicious inputs), and ultimately total failure. To combat this underperformance, there has been an introduction of companies offering Model Performance Monitoring (MPM) solutions, most of which require heavy integration, expensive in-house engineering resources, and man-in-the-middle oversight and remediation, all of which limits the scalability and productization of these MPM offerings. TRST is bringing a completely new approach to model monitoring that includes running a real-time audit of the data being ingested by the system to detect deficiencies and flaws, and then performing an auto-remediation or automated fix of that data to ensure continuous and optimal performance of the model. In addition, TRST is the first to market with an ROI generator, which is another reason why the MPM market is still in its nascent stages — enterprises need and want to understand the impact of a degrading model on their bottom-line in order to justify creating a budget line item for MPM spend.

MARKET APPLICATIONS
TRST is seeing significant applications in the financial services and insurance industries where there is not only the desire to more widely deploy Machine Learning, but also where the downside of unsupervised deployment can have a direct and negative impact on shareholder value. For example, banks and credit bureaus are using Artificial Intelligence and Machine Learning to process credit data as part of their mortgage approval processes. Because minority and low-income groups have less data in their credit histories, these models were making biased and inaccurate predictions on default rates. The same kind of faulty predictions are seen in the insurance industry where ML and AI are being leveraged for underwriting purposes — placing certain groups of people in high risk categories simply because the data on those groups were either insufficient or flawed. Other applications include Human Resources, where AI and ML recruiting platforms have shown to skew negative towards women, or Cyber Security where hackers can manipulate these ML and AI algorithms by inputting malicious data that masks anomalous activity on a network. Another interesting aspect to this market is regulation that is currently being explored here in the U.S. That regulation is centered around ensuring that these models come with enough supervision and oversight to ensure fair treatment of American consumers. If monitoring solutions are required for enterprises deploying AI and ML, that could cause the market opportunity here to grow exponentially.

MARKET ADOPTION
As a pre-seed company, TRST is still in the process of lining up initial design partners in the financial services and insurance industries.

TEAM

           

                   

                                                                                                                                                                                                                                               View fullsize                                                                                                          

         

   

 JANVEST VALUE-ADD
Janvest is the lead investor in TRST’s pre-seed round and as such has been instrumental in syndicating the remainder of the round - bringing in all of the other institutional investors to ensure a good mix of ‘smart’ capital around the table. In addition, Janvest has made a number of introductions to prospective design partners here in the U.S. including Greenlight Financial (Atlanta), a leading fintech company providing debit cards for children, SecureAuth (California), an innovator in identity protection, and Reveal Data (Chicago), which leverages AI and ML for eDiscovery. Janvest Partner Brian Rosenzweig will represent the fund on TRST’s Board of Directors.

JANVEST CONTACT INFORMATION
Brian Rosenzweig
Atlanta, GA
Brian@Janvest.com

Dafna Winocur Biran
Tel Aviv, Israel
Dafna@Janvest.com

Daniel Frankenstein
New York City, NY
Daniel@Janvest.com