>20k screenings per seconds*. Total transparency
*Benchmark achieved using commodity hardware on 100k synthetic dataset. Names in screenshot are illustrative.
.png)
Leverage real-time client and transaction screening with ultra-low latency, full configurability and explainability
Handles 20,000+ screening queries per second and scales horizontally – maintaining sub-5 ms latency* as you add nodes
More than 120 configurable parameters let you fine-tune the screening process to match your risk appetite
Combines diverse data sources and algorithms into a single, explainable match score– powered by deterministic matching for audit-ready decisions
Engineered from the ground up in Rust to deliver memory-safe, high performance screening at enterprise scale
Identifies and eliminates more than 20 distinct classes of false positives, saving you time and focusing your compliance review
Supports hundreds of languages via ultra-lightweight specialised “Nano Language Models”, to ensure accurate screening across multiple markets
Resolve identity across million of datapoints
Stream a name in any script and receive a normalized, similarity-ranked match list in milliseconds.
Phonetic, fuzzy, translation models drive industry-leading recall while keep false positives in check.
Submit client profiles and receive risk classifications in < 5 ms via deterministic matching against global watch-lists, PEPs, and adverse-media feeds.
Over 120 tunable parameters let you mirror your exact risk appetite, while built-in explainability makes every pass/fail decision audit-ready.
Reference-text screening for any event stream.
Feed short descriptors, counterparty fields, trade tickets, booking memos, payment reference, event strings
The API auto-extracts names and normalises them, and scans
20 000+ items per second while keeping latency under 5 ms.
Perfect for high-volume reference data where context is sparse but compliance can’t wait.
Replay historical cases, inject synthetic scenarios and benchmark your screening stack against external data and results
Run simulations and identify areas of weakness in your current processes
TECHNOLOGY
Means that we squeze every cpu cycle to sustain more than 20,000 checks/sec* with sub 5m latency can we make it "20,000 queries per second with sub 5ms latency
Minimise latency spikes and improve consistency in API calls leveraging Axum’s type-safe middleware friendly HTTP framework
We harden your screening stack against Common Vulnerabilities and Exposures. No unpredictable crashes during peak volumes
Our pipelines run on Tokio’s async runtime, spawning millions of lightweight tasks instead of heavyweight threads and scale linearly without context switch overhead that drags down a lot of the traditional tech stacks
We use multiple structures and algorithms to allow users to meet the most stringent SLAs on throughput and latency
.svg)
We combine a deterministic rule-based approach with machine learning to improve accuracy and precision of the queries
We are a team of compliance specialists, data scientists and system engineers united by a single goal: make risk decisions immediate, intelligible and dependable.
Years of building compliance applications for financial institutions, fintechs and corporate have taught us that traditional tools are slow, opaque and fragile. So we designed a new foundation, written in Rust, to deliver real-time performance without compromising rigour.
CORE TEAM
CEO and Founder
Head of Architecture
BOARD MEMEBERS
ADVISORY BOARD
CORE TEAM
CPTO
Senior Full Stack Engineer
BOARD MEMEBERS
BOARD MEMEBERS
.png)
Equip every compliance professional with the speed, clarity and assurance needed to protect their organisations and customers.
We believe that a compliance solution must be based on the client risk appetite and policies and not vice versa.
