Global Data Lab
Prospect 33’s Global Data Lab (GDL) is the leading R&D Lab for AI solution development in the wholesale capital markets. Our approach is rooted in our proprietary methodology, DaSMoDe, which emphasizes first principles thinking and careful data curation to ensure the reliability of our AI/ML models. DaSMoDe™ allows us to develop innovative solutions that drive genuine transformation in the industry.
We engage only the top young talent in the data science field, including Master’s Graduates and PhDs from renowned universities. We work closely with our clients to develop cutting-edge solutions that meet their unique business needs. Our typical project commences with an experimental phase, during which we use our first principles thinking approach to develop a proof of concept (PoC), followed by a commercial phase.
Our approach empowers financial institutions to access top talent, overcome challenges, and drive innovation in-house. By working with us, our clients have an invaluable platform for building and evaluating potential in-house expertise. At Prospect 33, we believe in the power of collaboration and are committed to helping our clients succeed. Our GDL provides clients and partners with access to the latest technology and expert talent, making it easier than ever to explore the potential of AI/ML in a real-world environment.
AI integration does not have to be another complex business process.
Contact us today to learn more about our unique methodology and how we can help unlock the full potential of your AI solution prototyping.
Data Scientists Available for Immediate Client Engagement
Credit Risk Early Warning Sentiment Analysis Tool
In collaboration with a leading US banking institution, Prospect 33 developed an AI tool that uses advanced Sentiment Analysis to provide Early Warning Indicators (EWI) to identify and manage credit risk. The EWI system utilizes natural language processing to analyze open-source news articles and extract risk-related information for a specific client or associate. This tool improves the efficiency and effectiveness of the client onboarding process and the client due diligence cycle, automating the manual review of news articles, reducing data redundancy, and minimizing human error.
At Prospect 33, we employ our proprietary methodology, DaSMoDe™, to train our AI models, leveraging a young, diverse, and dynamic workforce to build cutting-edge solutions that drive genuine transformation in the financial industry. Our GDL process enables our expert data scientists to work closely with clients to develop innovative solutions that solve specific business problems. We also employ first principles thinking and scientific principles throughout the process to ensure that we curate reliable data and deliver trustworthy production solutions.
Prospect 33’s AI-powered EWI tool is a valuable solution that can be deployed in client environments to enhance performance and reduce risk, ultimately helping global investment banks to avoid costly fines and penalties. Even though the tool is in its nascent stage, it has already demonstrated its effectiveness in identifying and managing risk for financial institutions.
Intelligent Transaction Monitoring Engine
Prospect 33 conducted a project to reduce fraud and money laundering following a documented money laundering event. Traditional rule-based systems and transaction analysis often fail to address the ever-changing tactics used by financial criminals. The proposed AI model takes a comprehensive approach to transactions, considering the attributes of all parties involved, which enables it to trace the origin and destination of illegally obtained assets more accurately. Utilizing advanced data engineering and graph technology, the model identifies relationships and exposes obscure connections between clients of varying risk levels. The resulting features are used to train a machine learning model for threat-level prediction. The model excludes personal identifying information to eliminate bias and comply with data protection regulations. The team leveraged scientific principles throughout the process, and the DaSMoDe™ methodology in GDL to ensure the reliability of the production solution. The model was tested on a holdout dataset and achieved an F1 score of 0.88, proving its reliability. Our young, diverse and dynamic workforce utilized first principles thinking to create this solution.
International Money Laundering and Petroleum Smuggling
Prospect 33 partnered with Intelligence Operatives to develop a suite of ML and AI tools aimed at optimizing the performance of wholesale capital markets, specifically focusing on fraud detection and transaction monitoring. The team utilized open-source data and technologies, such as graph networks and machine learning, to assist with the Know Your Customer (KYC) process, particularly in detecting criminal activities related to petroleum smuggling operations in the Middle East. The solution provides targeted recommendations, risk scores, and fraud detection by tracking connections to sanctioned entities and shell companies involved in illegal operations. Traditional fraud detection methods can be time-consuming and prone to human error, whereas AI algorithms can analyze vast amounts of data and identify patterns and anomalies that indicate fraudulent activity, allowing for faster and more accurate fraud detection. This solution represents a significant improvement in the efficiency, security, and reliability of wholesale capital markets. Prospect 33’s GDL team employed a graph database with relevant algorithms, fine-tuned with first principles thinking, to ensure maximum accuracy and reliability in the model’s outputs.
Workforce Capacity & Utilization Prediction
Together with a technology partner and leveraging DaSMoDe™ methodology, our diverse and dynamic team worked with the data of a significant US banking institution to develop a predictive modeling solution for utilization and capacity planning. By curating and analyzing data on tasks, vacation, seasonality, and employee attrition, we created a solution with 90% accuracy for a 50-day forward projection, suitable for departments with over 150 people.
Our GDL process and first principles thinking approach, as well as our commitment to scientific principles, allowed us to build a reliable production solution. Our team includes top young talent in data science and leverages our unique methodology to deliver innovative solutions that drive genuine transformation in the industry.