Advanced Analytics & Alert Management

Accuracy & Efficiency

OneTick Uses Advanced Techniques to Improve Alert Accuracy and Efficiency

Machine Learning (ML) and Artificial Intelligence (AI): OneTick leverages ML and AI to reduce false positives and trigger alerts more effectively. This includes self-tuning alert models that automatically adjust parameters based on market conditions, significantly reducing the need for manual tuning.

Alert Scoring with ML: Utilizes ML to score alerts, ranking them based on internal indicators of abusive behavior. This allows analysts to prioritize investigations by focusing on the strongest cases. For SaaS customers, this extends to unsupervised clustering to determine trader cohorts and compare behavior against norms.

Explainability: OneTick prioritizes white-box models for transparency and provides source code for all models (rules-based, statistical, non-linear). For complex LLM-based capabilities, OneTick ensures strict control over inputs and outputs, data access, and provides full traceability.

Dynamic Thresholds: Many alert models feature automatic dynamic adjustment or market-tuning of parameters based on real-time market conditions (e.g., spread, intraday moving windows) or historical profiles, ensuring alerts remain useful across diverse market scenarios without constant manual intervention.

Pattern Alerts: Allows for alerts to be defined on top of other exceptions or alerts, offering flexibility and potentially substituting for multiple overlapping surveillance perimeters for more economical runtime analysis..

Cross-Product & Cross-Market Surveillance

Detect Manipulation Across Correlated Instruments

More Interconnected? More Essential Than Ever: Cross-Product and Cross-Market

OneTick supports cross-product and cross-market monitoring based on both structural correlations (e.g., underlying assets and their derivatives) and non-structural correlations.

The system can ingest externally-supplied correlation matrices or OMD can generate them from market data as a 'correlations factory'. This allows for detecting manipulation across correlated instruments, even if not structurally linked (e.g., correlated chip stocks).

Examples include: "Mini manipulation" alerts (equity/futures underlier and options), Imbalanced Trading (ramping, momentum), and Insider Dealing across issuer products.

Efficient Alert Management & Compliance Workflows

From Detection to Resolution

Intuitive Investigation & Comprehensive Data Handling

The Data Capabilities You Require

Highly performant & customizable user interface with all of the data capabilities you require

Triage Dashboards: Offers web-based dashboards for efficient review, rapid triage, assigning, reviewing, annotating, and archiving alerts. Features include flexible filtering and grouping, contextual visualizations showing order states and market data, and multi-day/multi-quote views.

Order Book Replay (OBR): Provides a dynamic, visual representation of the order book, allowing users to step through interleaved market and firm/customer orders state by state. It can be launched directly from an alert, with the relevant instrument and time range pre-selected, and allows filtering by specific participants. News and alerts can be overlaid on the market timeline for enhanced context.

Smart Compliance Insights Lab (SC(A)IL): A SaaS environment based on JupyterLab for deep compliance investigations. It allows users to ask ad-hoc questions using SQL, Python & Pandas, or natural language (via an AI query assistant), create custom visualizations, and generate aggregate reports. It also supports creating "guest alerts" that integrate into the surveillance system.

Robust Data Ingestion: The OneTick engine supports the simultaneous collection of many real-time data sources, including FIX, Kafka, AMQP, and various proprietary feeds, with high-performance collectors tested at over 500,000 messages per second. It handles diverse data types like reference data, orders, trades, news, and corporate actions.

Data Quality Assurance (QA): The system performs dynamic data QA at several stages, including during loading and normalization, and prior to alert generation. It includes a large set of standard "sanity checks" to assess data quality and completeness, which can be configured to block downstream processing if issues are detected.

Robust Support & Implementation Partnership

Ongoing Support

The OneTick team is here to help you manage your projects with ongoing support.

Project Framework: OMD employs a hybrid project methodology combining elements of Waterfall, Scrum, and Kanban to ensure efficient implementation. There is a strong emphasis on a transparent project framework, providing customers with full access to project management tools like Jira and Confluence to track tasks, participate in planning, and monitor progress.

Comprehensive Support: Includes Product Support (updates, troubleshooting, issue resolution with defined SLAs), Managed Services (24x6 system monitoring and remediation for SaaS/PaaS customers), and Professional Services for customization work. Detailed escalation procedures are in place for complex issues.

Training: Offers various types of training, including business end-user training (triage, deeper investigation, CMS workflows), quantitative developer training (alerts development in Python/OneTick Query Language), and access to the OneTick Academy online learning platform

Would you like to see what we can do for you?

Empower your quantitative research and trading with OneTick's precision data & analytics today!