Bozena Adamczyk
Truffle Capital

For innovative companies in the financial sector, artificial intelligence is far from a novelty. Many solutions were initially designed with AI. Nevertheless, the arrival of generative AI marks the beginning of exponential development, creating new opportunities for tech industry players.

Artificial intelligence encompasses several forms of technology and applications.

Machine Learning (ML), which has greatly developed over the past few years, allows systems to be self-learning. Machines are able to identify patterns in data and make decisions with little or no human intervention.

Deep Learning,  a subcategory of machine learning, operates through neural networks. Inspired by biological neurons, these networks consist of layers of connected nodes. Neural networks, including deep architectures such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), are powerful tools in machine learning.

Decision Trees are an ML model used for classification and regression. They learn to make decisions by dividing data into subsets based on data features, forming a tree with decision nodes and leaves.

Finally, Natural Language Processing (NLP), also known as conversational AI, focuses on the interaction between computers and human language. It uses ML and deep learning techniques to understand, interpret, and manipulate human language.

Graph 1: The Different AI Algorithms
Source: Laurent Gimazane, Academic Division for Digital Education

AI, the Future of Finance

AI is currently in a phase of development and acceleration. Financial services has been the quickest sector to adopt these technologies. Indeed, over the past few years, banks have become digitalized and API-fied. They have revised their business models through strategic digitalization plans, transitioning from service companies to tech companies. According to a McKinsey Global Institute study, the use of AI to optimize fundamental banking operations and personalize services could generate value of more than$250 billion in the industry. All banks are now working on new use cases. For example, at BNP Paribas, the development and deployment of artificial intelligence are one of the levers of the ank’s 2025 strategic plan. Nearly a thousand use cases are currently in production, and the goal is to double this number and the associated value creation by 2025.

AI, a Source of Value Creation for Financial Services

GRAPH – Value Creation Enabled by AI in Financial Services

The entire banking industry is now being impacted by the deployment of AI. Each market type uses a technology (see graph). For instance, retail banking massively adopts generative AI, while risk and compliance departments utilize analytical AI and machine learning building blocks.

Une image contenant texte, capture d’écran, nombre, Police

Description générée automatiquement

Source: McKinsey – Dec 2023

Fintechs at the Heart of Transformation

Through their agility and capacity to innovate, financial technology companies are now at the heart of this transformation. They are a source of new solutions, and even systemic changes, for large groups.

Artificial Intelligence (AI) in Fintech Market size is expected to reach USD 17.0 billion by 2024 and is further anticipated to reach USD 70.1 billion by 2033 according to Dimension Market Research. The market is anticipated to register a CAGR of 17.0% from 2024 to 2033. 

The size of the AI market in fintech was valued at around  $8.23 billion to $10.1 billion in 2021.  It is projected to reach $61.30 billion by 2031, growing at a CAGR of 22.5% from 2022 to 2031, according to Allied Market Research

GRAPH: AI Market Evolution in Fintechs – Source: Dimension Market Research 

Fintechs have now largely embraced these technologies. It is already possible to distinguish five types of AI utilization by fintechs:

  1. Client Solvency Analysis

AI compares and analyzes data to determine if a client is eligible for the requested service. Thanks to a large language model (LLM), fintechs can interpret client data and assess risk factors. This instant solution  allows the emergence of new services, such as BNPL – Buy Now, Pay Later.

  1. Customer Support

NLP tasks include automatic translation, speech recognition, sentiment analysis, and automatic text summarization. Between 2022 and 2027, GlobalData predicts an average annual growth rate of 80% for the global generative AI market.

  1. Fraud Detection

Analytical tools powered by AI can collect and analyse vast amounts of data to learn user behavior patterns. This enables them to identify anomalies and detect warning signals indicating fraudulent activity. AI-powered LLMs allow fintechs to process sensitive or confidential information within their own infrastructure, thereby reducing the need to rely on external systems and minimizing privacy concerns.

  1. Process Automation

AI can take process automation a step further and enables increasingly complex tasks to be automated: document scanning, compliance verification, and document generation according to different formats and specifications. Several fintechs have embraced this technology. For instance, Traydstream has partnered with Infosys Finacle to create blockchain technology automating trade finance.

  1. Predictive Analysis and Decision Support

Marketplaces like foreign exchange have now largely embraced automated virtual assistants. Algorithmic trading enables the execution of the best possible transaction based on the data entrusted to the machine. These transactions are faster than those carried out by humans and provide a higher level of security. LLMs help analyze financial data and predict future investment trends, giving investors the confidence to make informed decisions to maximize returns. An LLM can undergo specialized training when intended for a specific domain or industry. Customization shapes the model to generate precise and relevant responses based on a domain, efficiently providing targeted information.

INSET: These Fintechs are at the Forefront of AI

Most fintechs use AI components in their solutions. Here are some examples of deployment:

An Asset for Fundraising

Generative AI is now a true asset for fundraising. Indeed, the tech companies that have held the largest fundraising rounds are those that place AI at the heart of their value creation.

Demonstrating that this topic is now essential, use of AIwas integrated this year into Truffle Capital’s Fintech 100, the annual ranking that analyzes the dynamics of the fintech ecosystem and the fundraising by the 100 most important fintech/insurtech players during the year. Nearly half of the respondents (49%) state that generative AI is authorized within their organization, through ChatGPT, while 12% of companies exclusively authorize the use of a generative AI developed internally.

Marketing (73% of respondents) and IT development and coding (53%) are at the forefront of use cases. Finally, 32% of respondents are convinced that the use of AI will have an impact on employee training, and 27% on the evolution of the business model of the company.

Case Studies

Case study 1: AI Revolutionizes Customer Relations

Customer relations is clearly identified as one of the fields of application of generative AI. Franz Fodéré founded Zaion.ai in 2017. This fintech aims to optimize the costs of processing customer interactions, particularly phone calls. Vocal interactions are, indeed, the most costly for companies, with an average cost per call of around €5. Thanks to its AI vocal solution dedicated to customer relations, Zaion automates the processing of simple requests, allowing advisors to focus on higher value-added tasks. Here is an interview with Franz Fodéré, president and founder of Zaion.ai.

Who are your clients and what are their requests?

Our clients are 80% financial services: banks, insurance companies, and mutual insurers. Banks and insurance companies operate by subscription, meaning that customer loyalty is key in their business development. This is why they care for their customer interactions and organize with internal call centers based in France. For financial services, customer service is at the heart of the activity. All players have now understood that AI would enable better and cheaper services.

What solutions do you offer?

Generative AI is revolutionizing customer relations by making “smarter” historical chatbots. Zaion relies exclusively on proprietary AI technology and offers, on the one hand, conversation automation solutions (called callbots), and on the other hand, solutions to support advisors and supervisors. To ensure the accuracy and reliability of the responses provided by our solutions, we fine-tune our AI algorithms on dialogues in French, using high-quality data from customer relations.

Generative AI brings more fluidity and personalization to conversations. Moreover, IAG automatically generates a summary of each call and saves precious time for the advisor who then only has to read and validate it. Finally, I can mention a critical use case for financial sector actors: the automatic evaluation of compliance or the quality of speech. By automating this task (historically performed manually), supervisors can focus on their expertise and support advisors in handling increasingly complex requests. At Zaion, these use cases rely on our LLM (large language model) called Vigogne, which, like all our AI technologies, is dedicated to customer relations.

What reception has your solution received during various funding rounds?

We have conducted two funding rounds for a total amount of €16 million, including €9 million in a Series A round in 2021. We are now preparing our third fundraising. I believe that solutions using AI are now the rare innovations that withstand the tech downturn. This is explained by the underlying growth rate. Over the next seven years, the annual growth rate of the AI market for conversational and vocal uses is estimated at 25%. The market in which Zaion operates is huge, enjoying very strong growth, and contains many irritants to address.

The bet on AI is therefore the least risky today, provided that the technologies are mastered to achieve the best level of performance.

Case study 2: Even Companies Seemingly Far from AI will be impacted

Launched in 2019, Obat supports construction companies in their administrative management. The platform offers a management software that allows for the creation of quotes, the issuance of invoices, and the sending of documents (purchases, expense reports) to accountants. An operational part has also been deployed, offering solutions for site planning and HR planning. The fintech, which has 15,000 business customers in total, now relies on AI to accelerate the deployment of new services. Here is an interview with Florent Liagre, founder and CEO of Obat.

When did you start using AI in your services?

AI is used for the improvement of our solutions. For a year now, we have significantly accelerated. We are in the process of structuring internally, creating a dedicated AI team composed of six people by the end of 2024, in tech, product, data, and operation. AI will allow us to make considerable advances, however, to engage in it, there is a significant technical cost. Therefore, it is essential to think carefully about investing without rushing.

What uses do you make of artificial intelligence?

We use AI on the product side and on the operational side. For example, in the management of purchases, we use OCR to extract the main data from documents. We are currently in beta testing for the launch of quote drafting from a voice note. This solution will allow a construction professional to dictate the elements necessary to draft the quote. The software will then search the client’s library for similar quotes to draft a new one. The document created is 80% finalized. The professional then only needs to verify and send it. This solution should be deployed very soon. In-house, we use solutions such as Dust, Modjo, GitHub Copilot, OpenAI, and others to accelerate the daily routines of our teams.

How do you highlight AI in your fundraising?

We held a Series A funding round of €12 million in November 2023. During this fundraising, we did not highlight the use of AI. We started implementing AI only a few months ago but believe it will significantly contribute to further amplifying our value proposition in the future: simplifying the daily routines and automating the processes of our clients. We continue to listen to them to identify use cases where we genuinely add value. These technological evolutions have the potential to accelerate the adoption of tools like Obat and the digitalization of certain markets.

Case study 3 : AI for streamlining exchanges 

Benjamin Rieder – CEO & Co-founder Levenue

What does Levenue offer? 

Levenue is a European wide marketplace where companies with recurring revenue are able to access affordable and non-dilutive financing to accelerate their company’s growth. Already operating in 17+ countries in Europe with the aim to provide founders with the financing alternative they have been waiting for.

The process is straightforward and eligibility is based on analysis of connected financial datasources: bank, subscription manager and accounting. 

When did you start using AI in your services?

The first AI we used was in 2023 when we started using AI translation when checking Articles of Association of our customers.

What uses do you make of artificial intelligence?

 We are using AI more and more : 

A big part of our customer succes work uses AI to accelerate onboarding by analysing company financial data, with customer interactions supported by AI chat bots

AI helps our analysts with translations, data cleaning and consolidation

AI helps us to ease the process of following up running financing contracts, by sending out automatic notification to internal (Levenue) and external (customers) involved parties, when our data analysis services see unexpected financial trends.