• Your Cart Is Empty!
Heliocor, 288 Bishopsgate, London, EC2M 4QP
The challenge for transaction monitoring systems is being able to keep up with an ever-changing landscape of trading patterns and products. Therefore, the ability to surveil across products and asset classes to identify malfeasance with a flexible, adaptable solution is key. 

Failure to deliver this flexibility can result in massive reputational damage and regulatory fines. 
Transaction monitoring in a buy-side or sell-side firm is a particularly challenging area of regulation for banks. 

The vast volume of data that need monitoring across asset classes, and the need to consider not just the transaction but all pricing and quote information, results in a data set that can no longer be processed using traditional databases and data processing techniques. 
The challenge for any senior compliance manager responsible for transaction monitoring is clear: to provide increasingly sophisticated, flexible and dynamic monitoring in real time, and across asset classes, all whilst reducing cost.

Robolitics™ is a very sophisticated real time monitoring platform. It uses basic rules to identify different types of reportable abuse. We keep these rules simple but then use A.I and Machine Learning to detect patterns, identify potential issues, and raise alerts  to the compliance team.

We also allow you to tune the sensitivity of these alerts to align them to the real risks in the business. This way, you can focus on the real isues rather than losing them in a myriad of false positives.

The “Below the Radar” items can then be surfaced to provide a 360 view of the client., raising an issue when an accumulation of minor infringements may merit a proper investigation.


Robolitics™ has been designed to meet this challenge by focusing on leveraging the best of the old and the best of the newand providing cross-asset class and cross-regulation monitoring of suspicious behaviours.  Robolitics™ has been designed from the bottom up to deliver a step-change improvement.  
Rule-based monitoring
We have a number of standard alerts based on typical risk behaviours. (Note: We can also create new alerts as new behaviour and risks become apparent through our RDK (Robot Development Kit)). These alerts can then be adjusted in real time to understand sensitivity of the rule; this we call “Machine Supported Learning”.  
Machine Learning
The Machine Learning model is a “trained module” that learns on the job. It provides the base functionality of alert reduction – e.g. not surfacing repetitive or repeat alerts.
Artificial Intelligence 
Robolitics™ reviews compliance officers' responses to alerts and learns on the job. It compares similar types of alerts either suggesting categorisation, making a recommendation for next best action, or creating a cluster of alerts so that multiple alerts can be managed faster.
Deep Learning 
Neural Networks / Deep Learning focuses on looking for patterns across the all transactions and or alerts. These often appear as previously unseen issues or coincidences.

These can then be surfaced to the compliance analyst rather than the compliance officer for investigation to identify if this is a real risk. This process is especially good at identifying new risks or uncovering unexpected patterns.
Robolitics™ is a monitoring, alerting, and reporting platform.  It contains a complex data model, over which multiple, easily tuneable “Robots” search to identify issues aligned to traditional models of “best practice” and then uses AI to further filter, cluster and categorise these events in order to remove false positives. 

Traditional solutions use cascading filters to monitor for behaviour whilst managing down a set of alerted transactions to a manageable set of alerts. Often, these rule sets become so complex and intertwined that no one understand how they work. They become fossil histories of risks the business has faced, rather than addressing current business threats.
Our solution originally based around the 24 reportable forms of market abuse described in MAD2/MAR provides real time monitoring across asset classes.  This set of rules can be extended using our RDK (Robot Development Kit) which simplifies the process of creating and tuning rules.
You can also tune these rules according to the nature of the your business and market conditions to deliver risk based monitoring.

The Robots then “learns on the job,” looking for transaction patterns, surfacing hidden relationships, linking alerts between parties and understanding client activity to improve the identification of real risks.
You can quickly create a new alert based on suspicions or identified behaviours of either clients or staff.  As a result, you can be sure that repeat behaviour is identified and addressed quickly. Over time these alerts mature as the AI identifies and removes repetitive tasks.

Robolitics™ also cuts investigations time on alerts by
  • Surfacing a comprehensive data set – covering the transaction, the market, and historical behaviour
  • Clustering similar issues – using AI to group alerts allowing them to be handled more efficiently 
  • Proposing best action  based on previous responses to similar alerts
  • Delivering customer sensitive compliance - enabling you to tune alert sensitivity to the client risk profile

Customer Onboarding with Dokstor

Improved KYC and On-Boarding processes under AML 4 & 5. Dokstor ensures a common and consistent approach to KYC.


Real-time monitoring of inbound and outbound payments and transactions for retail, wealth, and fintech businesses.

Technology Overview

Robolitics is available as a Cloudera Stack based on Hadoop for high volume and real time mission critical solution with hot failover. A lightweight SQL option is available for smaller enterprises.