OpenTalks.AI - Žygimantas Medelis, Polyglot data interrogation for compliance

Transcript

1. ‘INTELLIGENCE OF EVERYTHING'™
2. POLYGLOT DATA INTERROGATION FOR COMPLIANCE
3. COMPLIENCE CHALLENGES KYC is long, costly & labor-intensive Volume is growing substantially Ever increasing regulation False positives reduce productivity False negatives are the hidden danger Banks work with limited information –watch lists, prior transaction patterns, ADVERSE NEWS Compliance doesn’t increase competitive market advantage Lack a single view of the legal entity Multiple fragmented pin point solution Static view of risk COST VOLUME REGULATION FALSE POSITIVE/ NEGATIVE REACH ADVANTAGE VIEW COMPLEXITY RELEVANCE
4. COMPLIENCE DATA: HOW TO MAKE SENSE OF IT?
5. ELEMENT OF COMPLIANCE MANAGEMENT INVESTIGATION DETECTION - MONITORING DETECTION - SCREENING ENRICHMENT & VALIDATION ON BOARDING (Dynamic Question) DYNAMIC ONBOARDING QUESTIONNAIRE ENTITY CASE CREATION REALTIME PROCESSING PERIODIC PROCESSING INVESTIGATION REPORTING & ADHERENCE Identity questions Business classification questions Financial activity questions Search indexes and internal for correctness & completeness Asses Direct and Indirect Risk Score Construct behavioural patterns (ML rules) Ongoing customer screening Reputation Monitoring Transaction Screening Detect changes in key entities Reputation Monitoring ML rules enforcement Analysis tools Analysis tools Analysis tools Reporting & SAR Creation Reporting & SAR Creation Reporting & SAR Creation SUMMARY RISK ASSESSED ENTITY CASE INVESTIGATION CREATION SAR DISSEMINATION ID&V KYC TM OUTPUTS
6. ELEMENT ENTERPRISE GRADE, AI POWERED PLATFORM FOR OPERATIONAL EFFICIENCY AND ADVANCED DECISIONING ENGAGE VISUALIZE DISCOVER DATA FLOW OUTPUT ‣ Easy to capture multi-format Multilanguage data including natural language (unstructured), Real Time and data in rest ‣ Triaging, extraction, data cleansing, data quality validation & transformation ‣ Prepares data for harvesting / storage ‣ Classification engine (ML) for new data frontiers and context BIG DATA COLLECTION IN CONTEXT STRUCTURED, UNSTRUCTURED, ENRICHED DATA PREDICTIONS NETWORK ANALYSIS, M/D LEARNING KNOWLEDGE GRAPH ENTITY MGMT CONSTRUCTING FULL KNOWLEDGE GRAPH UNIVERSE. ANALYSE RECOMMEND AND PREDICT ‣ Extracts entities & relationships in line with the specific ontologies, builds clusters, self-emerging networks ‣ Rules and algorithm to resolve, match and disambiguate entities ‣ Machine/Deep learning to automatically build knowledge and further enrich entities in all available sources ‣ Analyze Risk Recommend opportunity/ predict behaviour ‣ Engage and analyze ‣ Self generated alerts ‣ Intuitive query ‣ Make and test hypotheses ‣ Time line specific behaviour and movements ‣ Predictive analysis VISUAL INTERACTION INTERFACE FOR ANALYSTS DISCOVERY AND DATA VISUALIZATION ‣ Manage the entire Drag & drop visual interface for defining data pipelines (workflow) ‣ Build manage and refine policy enforcement (rules) and business logic ‣ A range of different operators to manage sources, operation, storage search, GIS etc ‣ Build, execute, publish, package, test and refine the entire Machine Learning, Deep Learning, AI requirements ‣ Batch, streaming & ML/DL workloads support ‣ REST API’s for third party integration ‣ Scalable & handles massive volume of data OWN/CONTROL/MANAGE PERSONALISE YOUR ASSET
7. UNPRECEDENTED TRANSFORMATION IN THE WAY KYC PROFILES ARE CREATED AND USED ‣ Auto ingestion of internal KYC profiles and documents and further enrichment from external sources ‣ Extraction of entities & relationships with timeline view ‣ Self generated alerts to highlight changes in KYC, hot topics etc ‣ Identification of associated entities and build clusters and networks, including disambiguation of entities ‣ Enriched overview of customer – CIP and CDD info etc. ‣ Ability to further drill down into individual entities and clusters ‣ Deep dive into leadership team with a timeline view to identify controllers of the company ‣ Automated financial information gathering ‣ Specific module focused on risk alerts for KYC Profile – Sanctions, PEP, Adverse News, Fraud, Cyber, Blacklists, Criminal, Dark web activity etc ‣ Rich visual interaction interface for analysts to get snapshot of all latest news, press releases, adverse news, including auto tagging of key words ‣ Rich analysis of social media profile ‣ End to end system capable of seamlessly integrating KYC intelligence to solve transaction monitoring alerts ‣ Ability to create KYC profiles with basic seed data ‣ Risk assessment of the customer by considering rich set of parameters defined
8. ELEMENT THE ART OF IMPOSSIBLE: UNPRECEDENTED TRANSFORMATION IN THE WAY BUSINESSES ARE CONCEIVED AND RUN DATA FINDS DATA ENTITY ANALYTICS SELF GENERATING TARGETS DYNAMIC PROFILE AND RISK SCORING ELASTIC DATA MODEL Unique algorithm for entity enrichment allowing the automatic and dynamic process of building a knowledge tree on a specific topic or entity: oactual ontology which describes entities, properties and relationships Ultimate entity analytics engine based on our “intelligence” experience: oDeeply understand the nature of avatars and identity hiding oA unique, one of its kind functionality - Entity Match, Consolidate, Disambiguate The next generation of Cognitive Computing is here: oElement infers, connects and reveals even before one could raise the question! oProactive “connecting of the dots” Constantly gather information on the relevant entities and the network oFrom multiple data sources – either internal, external or non traditional (OSINT) enabling a dynamic profile and risk scoring We don’t believe in imposing ourselves on the client; nor do we wish to assume how its operation will evolve. We therefore developed an Elastic Data layer allowing our partners the freedom and flexibility to shape their future
9. FINANCIAL SERVICES TOP GLOBAL BANK SUFFERS FROM EVER INCREASING COMPLIANCE COSTS ‣ Global bank (+100BL) Top 15 ‣ Over 4M Corp Customers ‣ Compliance cost reaching 7% of expenditure (3K per OB) ‣ Extremely complex technology to keep regulation in place ‣ Both compliance and FIU in a single Unit 1 . ‣ KYC is long, costly & labor-intensive ‣ Volume is growing substantially ‣ Ever increasing regulation ‣ False positives reduce productivity ‣ False negatives are the hidden danger ‣ Banks work with limited information –watch lists, prior transaction patterns 2. 3. ‣ Reduced operational cost by 50% ‣ Increased Analysts productivity by 100% ‣ Cost per customer OB brought to less than $150 ‣ Single system to manage the entire compliance and FIU spectrum 4.
10. FINANCIAL SERVICES TOP 5 GLOBAL BANK STRATEGIC INITIATIVE ‣ Global bank (+230K Employees) ‣ Operates in over 70 countries ‣ +35M global customers ‣ Extremely complex technology to keep regulation in place 1 . ‣ Ever increasing operational costs due to size and distribution ‣ Very hard push to break it to 5 regional banks ‣ Leverage the bank global presence ‣ CEO strategic initiative 2. ‣ Building a massive knowledge graph of business , ownerships, directorships etc ‣ Automatically building a real time and dynamic knowledge on customers from X00,000s number of sources including new sites, idexes etc oraganising it in an actionable manner ‣ Enabling X,000s of relationship managers to build tactical approach to generate new business 3.
11. FINANCIAL SERVICES SINGLE GATEWAY TO MARKET INTELLIGENCE ‣ Big 4 ‣ Largest in Financial Services ‣ Advising 100% of fortune 1000 1 . ‣ Too many reports which are either sectorial or regional ‣ 100% are historical ‣ No differentiation ‣ Missing pockets of opportunities 2. ‣ Digest, analyse and visualize data structured, unstructured real time and in rest ‣ Define and execute value models ‣ Segmentation as option ‣ Ad hoc campaigns 3.
12. MICHAEL@BLACKSWAN-TECHNOLOGIES.COM CONTACT US NEW YORK, US TEL AVIV, ILS WWW.BLACKSWANTECHNOLOGIES.AI TEAM@CYBERTEKLAB.COM MOSCOW, RUS WWW.CYBERTEKLAB.COM