14 accelerator applications use advanced AI techniques to address complex business challenges like fraud, margin call management and LCR calculations
Synechron Inc., a global financial services consulting and technology services provider, has today announced the launch of “Neo,” a set of Artificial Intelligence (AI)-based solutions for the financial services industry. Neo uniquely brings together Synechron’s digital, business and technology consulting to allow financial institutions to deploy cutting-edge, AI solutions that solve complex business challenges.
Synechron has built 14 reusable applications – Accelerators – that allow financial institutions to reduce time-to-market when applying AI to enhance business operations, reduce operating costs, and create better client experiences. Synechron’s AI Accelerators applications have taken a business challenge that can be best solved through AI and re-architected business processes to build a solution powered by artificial intelligence and optimized for user experience. The Accelerators use techniques like Natural Language Processing (NLP), Chatbots, Robotic Process Automation (RPA), Cognitive Machine Learning, Data Science, and Robo-Advisors to address a range of use cases. In addition to custom development work, Synechron has partnered with specialist FinTech firms such as Quantexa, Yseop and SQREEM.
Faisal Husain, Synechron Co-founder & CEO, said:
“Financial institutions are looking to implement the latest technology to address real-world problems in financial services. Neo and Synechron’s AI Accelerators will be pivotal in helping clients be at the forefront of technological advancement, while providing a comprehensive set of tools to ease and streamline processes. This will allow businesses to deploy technology-enabled processes that augment the role of individuals, allowing them to be elevated to higher-value business tasks.”
The AI Accelerators broken down by the core underlying AI technology include:
- Natural Language Processing (NLP) and Generation (NLG)
- Automated Data Extraction uses NLP to achieve automated data extraction and intent realization, allowing firms to pull data from earnings reports and other sources and contextualize its intent.
- Automated Financial Advice Generation can be achieved by using NLP to extract CRM data and NLG to reach a compliant conclusion through real-time queries and contextual user information.
- Automated Executive Summaries written in plain language using NLP and NLOG.
- Chatbotsserve as either an internal virtual assistant or a frontline customer representative and have been created with an understanding of financial services business operations and systems integration expertise. The accelerators include: BankBOTfor personal banking,TraderBOTfor traders,LoanBOTfor mortgages, andInsureBOTfor insurance.
- Robotic Process Automation (RPA)
- Client Onboarding uses OCR + NLP to pull information from images of documents such as driver’s licenses and passports to auto-populate forms and create a frictionless onboarding experience.
- Automated Resolution of Failures “Breaks” in Reconciliation processescompletely automates GL reconciliations.
- Automated Margin Call Management analyzes emails using NLP and automatically understands relevant margin call information based on pre-set criteria.
- Automated Pitchbook Generation allows financial institutions to automatically generate presentation decks by understanding what content is on the slides and the appropriate disclosures required based on that information.
- Cognitive Machine Learning
- OTCPrice Automation Synechron is using machine learning to derive real-time OTC pricing for illiquid OTC products where this data is currently decentralized and difficult for traders to factor into their pricing models and further using that data to advance collateral management reporting.
- LCR Reporting uses historic data and machine learning to come up with a reliable intraday liquidity estimates for LCR reporting.
- Data Science
- Customer Insights has four Modules for Banks, Credit Cards, eCommerce and Mortgages that allow banks to bring together their Know-Your-Customer (KYC), Banking and Credit Card Data into a database, and join them with the customer’s online behavior (if opts in) via web and social platforms.
- Product Recommendation uses behavioral analysis to understand customer patterns for new client acquisition.
- AML/Fraud Detection uses AI and behavioral analysis to identify potentially suspicious activity indicative of money laundering and fraud.
- Artificial Intelligence Accelerator for Robo-Advisors allows wealth managers to create a hybrid robo-advisor that augments their existing services with an automated platform, creating the self-service experience clients are looking for, balanced with the high-touch, high-trust experience advisors are known to deliver with added capabilities like social investing, chat and more.
The AI Accelerators apply some of the most advanced AI and Deep Learning techniques, and algorithms including Google Tesseract’s OCR and binarization, the NLTK platform for sentiment analysis, SARIMA for time-series data, Self-Org Maps for clustering, Deep Learning Recurring Neural Networks (RNN) for dynamic classification, Support Vector Machines for classification, R for statistical modeling and reinforced learning, multi-level Perception for classification, and Apache Spark for big data and machine learning. They also draw on deep financial services expertise, and foundational programming languages and tools including OpenNLP, R, Python, Jflex, PDFBox, MS LUIS, Tessaract, Spark, Apache Solr and mongoDB, among others.
The AI engine, business and technical analysis at the core of these accelerators can be applied to additional use cases to progress more quickly with similar initiatives. Along with the use of the AI applications, the Accelerator Program offers access to Synechron’s team of consultants, technologists and digital teams who are experts in financial services business processes, products, regulation, operating models and data architectures which are critical to constructing affective AI applications.