andre-donaher

Andrew Donaher

Vice-President, Consulting Expert

juryn

Chris Juryn

Vice-President, Emerging technologies

navjeet

Navjeet Nehra

VP Consulting Services, Toronto Financial Services

The unprecedented power of Artificial Intelligence is in sharp focus for organizations everywhere today and CGI’s AI experts are providing market-leading guidance on the critical need for responsible AI adoption, while delivering a sustainable competitive advantage as the rapid pace of change accelerates.

While AI is already enhancing productivity in health, manufacturing, energy, insurance and beyond, our AI for speed to trusted action series explores emerging challenges and solutions that demand attention for organizations in different sectors – beginning with a detailed look at finance and banking. Our series features insightful use cases among today’s future-focused organizations.

Future-focused businesses realize that AI has the potential to redefine their processes, operational efficiency and customer experiences for a new era of competitiveness. In today’s hypercompetitive environment, leading businesses are also wisely recognizing that there is no time to lose on AI adoption. They simply cannot risk being left behind.

The financial services sector was estimated to spend more than US$10 billion in 2023 to improve customer experiences, automate processes and accelerate back-office functions such as threat detection and fraud analysis, according to IDC Research. And as businesses race to innovate, transform and compete in the digital economy, AI will play an increasingly pivotal role.

While there’s no question that data makes the world go round in the digital economy, financial organizations are being overwhelmed by a relentless and fast-growing volume of operational and customer data. Beyond the pressing need to transform data management and accelerate decision-making, finance-sector businesses are also struggling to acquire modern skills that are now critical to innovation – from data scientists and data engineers to professionals who understand the implications and proper use of AI.

Also dominating the business agenda is the need to hyper-personalize experiences, combat fraud, improve data security and optimize operational performance, governance and risk mitigation. And not to be underestimated is the reality that finance organizations are tackling significant challenges within a troubling global environment that includes recession fears and economic uncertainty, geopolitical volatility and conflict, rising inflation, supply-chain disruption and evolving regulatory scrutiny.

Generative AI is boosting performance, efficiency and revenue
Since exploding onto the scene, Generative AI continues to enhance operational performance, efficiency and revenue generation in the fast-evolving finance sector. While it is transforming organizations in unprecedented ways, it is by no means a panacea. Successful implementation requires solid data, processes and knowledge of its capabilities – and limitations – to meet expectations and drive your business forward. Here is a quick overview of the impact that Generative AI is having in the finance sector today: 

Hyper-personalization
Personalized service has long been a hallmark of the customer experience for financial institutions and in today’s consumer-centric economy, a personal touch is indispensable. CGI’s unique Customer Lifetime Value approach – focused on the pivotal need to Know Your Customer, Know Your Products, Match Together, and Optimize – has made a difference for clients around the world. And now, the remarkable power of Generative AI is enabling us to mine those customer-value insights in new ways to create hyper-personalized experiences and micro-directional customer support. From establishing whether a service or product fits a customer’s specific needs to contextualizing support and services in real time, Generative AI is changing the rules of the service game in the race to enhance customer loyalty and growth.

Operational efficiency
According to various sources, 90% of the world’s data has been created in just the last two years and financial institutions have been among the trailblazers in the creation and consumption of data in the modern economy. At the same time, ask any financial analyst and they will tell you it is virtually impossible to keep up with the endless waves of data now inundating their organization around the clock.

Generative AI is ‘turning the tide’ – revolutionizing data management with its unprecedented capability to compile and analyze data and deliver timely insights for smart decision-making by employees and clients alike. Sector leaders are implementing data lineage to ensure data is accurate, trustworthy and used appropriately, ultimately improving decision-making while minimizing risk and mitigating ‘AI hallucinations’ – information that is non-existent or imperceptible to human observation. At the same time, it is worth noting a CGI core belief that we emphasize with clients: while AI makes experts more efficient, it doesn’t make everyone an expert – meaning you always need a human in the loop.

Legacy-platform modernization
Financial institutions have been in the data and analytics game for a long time, and many are reliant on legacy platforms that, while critical to effective data management, have limited capabilities to evolve. Also limited today, amid changing demographics and an aging workforce, are the human skills needed for COBOL or Job Language Control (JCL) programming. The result? An array of new risks are emerging amid the pressing need to manage and maintain these systems, ensure they stay up, and deliver new functionality at unprecedented scale and speed.

To mitigate these risks, CGI experts have been helping organizations tap into the power of Generative AI. We are showing how it can interrogate applications and read code to: understand what is happening – the data lineage and business rules built into the system; populate requirements documents; and create a new code-base in a new platform or language. The code, while at times imperfect, helps expert coders and analysts increase productivity and accelerate delivery time. And keep in mind that you always need a human in the loop.

GenAI real-time analytics
The key to effective analytics is an approach that focuses on speed to trusted action. But beware – this doesn’t simply mean allowing real-time data to stream into your system from every available source. It means receiving precise and informative answers to your questions, accompanied by instant visibility into the calculations made and data used – rather than waiting weeks or months for ‘a new report.’ 

CGI has been working closely with organizations to implement Generative AI and build real-time analytics engines that sit on top of existing data platforms and repositories. These transformational engines build queries on your data in real-time, consolidate the data, explain the rules applied, and apply data-visualization best-practices to expedite analysis. Generative AI renders obsolete traditional methods in which questions posed typically generate more questions until a final answer reached. It enables speed to trusted action – delivering critical business insights faster, while providing the transparency and lineage needed to understand, trust and rapidly respond to the results generated.

Building on a foundation of success with AI
CGI’s speed to trusted action approach aligns and operationalizes responsible AI for a sustainable competitive advantage across a growing array of use cases that include:

  • Operational efficiency and cost optimization; enhanced risk management, fraud detection and compliance;
  • Personalizing customer experiences to grow revenue and enhance customer loyalty;
  • AI platform modernization to remain future-proof;
  • Responsible AI use – ensuring ethical, unbiased data use to maintain customer trust.

Combatting fraud as threats soar
Fraud and customer data breaches continue to rise and are top-of-mind for every customer-facing organization today. In our work with a leading Canadian bank, the organization’s fraud team identified the need to enhance its existing IT architecture and the ML technology stack, and transition to ‘advanced machine-learning’ techniques such as unsupervised learning, deep learning and more. The client faced difficulty kickstarting advanced methods of machine learning model development and production, as well as a lack of collaboration between its data scientists and rest of the organization.

The goal was to scale existing models and increase fraud-detection accuracy. Our AI experts conducted strategic alignment workshops and engaged in interactive discovery sessions to establish the required future state while assessing key impediments to progress that needed to be addressed.

By developing a tailored questionnaire and analysis, we identified the organization’s priorities and co-created the final-state solution needed to achieve the client’s targeted outcomes. Our detailed plan allowed the client to prioritize the required work, link that work back to tangible and measurable outcomes, and develop the case for change. CGI’s implementation featured a comprehensive ‘future state roadmap and playbook’ that included:

  • In-depth assessment of the client’s current state;
  • Analysis of all ML and supporting tools within the future-state architecture;
  • Visual architecture of proposed tools, associated integration and potential ML outputs;
  • A customized mobilization plan to actualize the future state that includes operational and technical imperatives.

The bank’s fraud detection is now up to speed in responding to today’s threats and protecting customer data and privacy.

Taking quality control to a new level
Our AI specialists also collaborated with a major global asset-management firm facing significant quality-control challenges. The Canadian-based client was losing millions of dollars annually amid inconsistencies in creating and monitoring new customer accounts.
The quality-control function chose accounts at random to monitor and lacked the intelligence to identify customer accounts most likely to have issues demanding rapid solutions. Key customer data was spread across disparate sources and the client also possessed an imbalanced dataset for its model training.

CGI experts helped migrate the firm’s legacy system to a modern microservices architecture and a highly scalable cloud platform. We also implemented an ML-based ‘Anomaly Detection’ solution to identify accounts most likely to meet the client’s quality standards. Our ML-driven solution has significantly reduced costs and improved the customer-onboarding experience. 

Optimizing collections outcomes
Rapid and precise communication with customers is essential to success in today’s customer-centric digital economy. Unfortunately, a leading US banking client endured serious challenges amid the need to implement an AI model that would instantly predict the best channel and time to communicate with customers facing collections. 
Our AI experts collected a dataset among 100,000 accounts possessing several key attributes such as: the customer’s profile, the history of customer communications with the institution, the customer’s social-network activity, and their past collections activity.

We utilized open-source solutions to develop a series of machine-learning algorithms that revealed the best customer communication channels – whether cell or home phone, email or text – and the ideal time of day to connect. The bank’s customer-collections function has reached a new level of speed and customer satisfaction thanks to the power of AI and ML.

Matching customer and third-party financial data
A major US banking client needed a solution to match customer data with important data and information being generated by outside financial sources such as S&P Market Intelligence, Dun and Bradstreet and Moody’s. The goal was to enrich customer information with timely and informative financial statements, stock-performance reports and more.

Our experts introduced an automated approach that reduces human intervention and manual activities – automatically matching customers with relevant third-party data they could use and ultimately enhancing service, while also assisting in compliance reporting. We applied AI to build a ‘record-linkage framework’ – connecting the bank’s internal customer data with relevant third-party data. The matching framework can be configured based on specific customer requirements and it features an automated review process that continually tracks and improves the AI model’s accuracy and effectiveness.

How to succeed: A sustained approach is crucial
The responsible AI journey is never complete. Forward-looking organizations are wisely building skilled teams and capabilities to deliver ongoing progress on the AI front based on existing proof of concepts. CGI’s strategic, four-stage roadmap to effective AI adoption is making a difference for clients on the AI journey:

  • Strategize: CGI’s AI experts conduct strategic alignment workshops to understand your initial objectives for AI. We conduct a competitive scan on AI use and investments to date, then explore existing product lines to identify potential AI and ML use cases.
  • Prioritize: ML use cases are evaluated and scored against a well-defined framework. Collaborative and iterative sessions are held with technology leads and product or business owners on evaluation results.
  • Build: Our experts rapidly prototype shortlisted ML use cases by building a low-cost AI model on a subset of actual company data. We quickly deliver a basic computing infrastructure and environment for the AI model’s operation, and identify any correlations in data, gaps or dependencies.
  • Mobilize: A solid mobilization plan to scale feasible ML use cases is provided using CGI’s proven expertise and best practices. We then connect with leadership, business owners and technical owners to discuss proof-of-value results and ensure alignment with the client roadmap.

We also stress to clients starting out on the AI journey the importance of having the right operating model in place for successful AI adoption - one that aligns with their specific and unique business needs, including appropriate capabilities, skills, teams and roles. As success stories continue to emerge, it’s imperative to ensure that responsible AI adoption is a key priority in every use case. Recent market research indicates that consumers are cautious about the growing use of AI in various aspects of business, with 75% of consumers citing concerns about AI’s potential for misinformation.

AI requires immediate and continued investment and innovation as models evolve and proliferate from month to month. Simply put, you need to invest in AI, not only today, but on a sustained basis with a strategic long-term goal. As we continually stress to clients: ‘avoid paralysis by analysis.’ There is no time to lose in today’s dynamic reality.

CGI takes a scientific approach to evaluating AI and ensuring responsible, trustworthy and secure AI adoption. Our proven methodology is industry-specific – we focus on the unique needs of each sector and the businesses operating within it. Data use needs to be ethical and unbiased and data privacy must be adequately protected. Security and accountability need to be ‘baked into’ the AI model. Our approach includes an AI Code of Conduct to ensure that the principles of transparency, bias, privacy, security and moral responsibility are met throughout the AI engagement.

Our experts validate the tailored AI strategy and each use case to ensure value from the AI model. Our tailored roadmap to implementation focuses on people, processes and tools. Crucial to success is the need to optimize human decision-making during AI use – ultimately aligning human skills and decision-making responsibilities with AI-generated insights and outputs. AI’s data-based suggestions and outputs need to be evaluated by humans to determine their accuracy and whether they should be acted upon or rejected.

CGI is a recognized industry leader in AI services
CGI was named a leader in Canadian AI services by IDC MarketScape in its Canadian AI Services 2022 Vendor Assessment. The report, a first for the Canadian market, identified three of the most-critical services that businesses are seeking from AI providers today: cybersecurity, intelligent automation and industry-specific AI solutions. CGI’s noted ‘strengths’ include the ‘breadth and depth’ of our:

  • AI services capabilities;
  • Tools to identify use cases for customer AI projects;
  • Tools to deliver AI services;
  • IP, methodologies and tools to manage customer data, analytics and AI maturity.

IDC also highlighted CGI’s industry expertise in Canadian analytics and AI implementations in several key sectors – including financial services. As financial organizations increasingly embrace the power of AI, CGI is ambitiously building on our trusted AI foundation to accelerate their AI journey. As announced in July 2023, CGI will invest $1 billion over the next three years to support continued expansion of our AI services and solutions.

“We believe that we are at the beginning of a new wave of innovation and that the business value of AI will be achieved through the combination of human expertise and ethical use of technology,” says CGI President and CEO George D. Schindler. “Today, our consultants are drawing on proven AI use cases and pre-built, industry-focused solutions powered by trusted domain data sets to help clients navigate the hype and make the best return on investments.”  

About these authors

andre-donaher

Andrew Donaher

Vice-President, Consulting Expert

Andy is a data, analytics and digital expert who works with leading clients throughout North America to convert digital and analytics capabilities into actionable insights that drive competitive advantage. He has a long and successful track record of building, leading and executing global, enterprise-scale transformational ...

juryn

Chris Juryn

Vice-President, Emerging technologies

Chris is an experienced leader with expertise across the tech industry leveraging a focus on delivering business-focused outcomes. He’s held technical roles across all aspects of the SDLC with positions in QA, Performance Optimization, Project Management and Development. ...

navjeet

Navjeet Nehra

VP Consulting Services, Toronto Financial Services