2023 Integrated Annual Report

Introduction

THREE KEY TRENDS FOR CROSS-INDUSTRY IMPACT
  • Multimodality: we are seeing an expansion beyond text to include images, audio and video in generative AI applications, revealing untapped potential.
  • Small is beautiful: there is a shift towards smaller, cost-efficient AI models that maintain performance, addressing scalability and affordability.
  • Beyond technology: generative AI isn’t just a technological trend enabling the development of transformative applications. It’s also about reimagining company value chains, processes and organizational structures to unlock new revenue streams and push new frontiers in operational efficiency.
A BROAD PORTFOLIO OF BESPOKE GENERATIVE AI SOLUTIONS

Our AI-generating solutions have evolved to span transformation, solutions and operations. Designed to deliver value and improve operational efficiency, they include:

  • Generative AI strategy enables C-suite executives to define and prioritize the most relevant generative AI use cases for their business, demonstrate the tangible value that can be achieved, and lay the right foundations in terms of people, process and technology for scaling their generative AI investments while mitigating the risks.
  • Generative AI for Customer Experience enhances customer experience with generative AI front-end applications and dedicated generative AI assistants across customer journeys, sales, service, commerce and marketing. It allows hyper-personalized customer experience with synthetic design assistants, elevates customer self-service with personalized AI conversational assistants, augments customer care services with task and knowledge assistants and boosts the performance of sales teams with a product and offers knowledge assistant.
  • Generative AI for Software Engineering helps improve efficiency and quality across the whole software life cycle (from design and coding to documentation, testing, deployment, and operations), accelerate the time to market for new software, and reduce the technical debt of businesses by facilitating large modernization programs of legacy software. It also enables increased security with a reduced attack surface by automatically identifying bugs or vulnerabilities and proposing adjustments to software development teams.
  • Custom generative AI for Enterprise enables businesses who have proprietary data to develop custom generative AI assistants, using prompt engineering, retrieval augmented generation (RAG) and/or fine-tuning. We have designed a Reliable AI Solution Engineering (RAISE) platform to deploy generative AI at scale. It includes an experimentation module to enable validation of tech choices versus performance in the early stages of development, as well as guardrails, data platform automation and FM (Foundation Models) Ops. These customized models, building out from company know-how, enable our clients to generate new revenues and significantly improve their operational efficiency.

LEXICON

LLM

Large Language Models use advanced machine learning techniques, in particular deep learning, to decipher and master human language.

NEURAL NETWORKS

A model used to teach computers, inspired by the neuronal organization found in biological neural networks.

PROMPT

Any written command sent to an ‘artificial intelligence’ specializing in the generation of content, such as text or images.

OPEN SOURCE AI

Designed to be accessible to the public: anyone can view, modify and distribute the code as they see fit.