Blog
Category

How AI changes wealth & asset management

In the field of wealth management, Etops is committed to innovation and keeping pace with the latest developments in tech and AI. We are dedicated to meeting the evolving needs of our clients and continually improving our products for higher efficiency and better client value. The Etops Data Science team is pleased to introduce “Etops Co-Pilot”: a chatbot engineered for wealth managers, family offices and asset managers. This AI-driven product integrates CRM and portfolio analysis with real-time data, creating financial insights with ease and precision.
Santiago
CPO - Etops
By
This is some text inside of a div block.
March 20, 2024
5 min read
Share this post
Retrieval-Augmented Generation: A personal assistant that actually helps with your work

In the world of AI and Natural Language Processing (NLP), Retrieval-Augmented Generation (RAG) is a concept that is attracting interest. This innovative method blends the retrieval of data with the promise of Generative AI models, such as the transformer-based ChatGPT model, crafting a system that operates as your personal assistant without the complexities of training an extensive language model.

Although it is still early, Co-Pilot has shown promising capabilities. From acquiring real-time stock prices to synthesising summaries of manager-client interactions, and from fielding intricate portfolio inquiries to recapping market trends and recommending potential research on markets and products, Co-Pilot equips Wealth Managers with the tools to access deep web information swiftly, enabling them to make well-informed decisions with confidence.

The user query is sent to the LLM and depending on the interpretation by it, we fetch the relevant source (e.g. web or internal API)

Deciphering Retrieval-Augmented Generation aka RAG

Retrieval-Augmented Generation marries two core AI methodologies: retrieval models and generative models. Retrieval models are proficient in sifting through datasets to locate pertinent information, whereas generative models excel at fabricating new content that is contextually coherent. RAG encapsulates these capabilities, fostering a symbiotic dynamic that augments both aspects of the technology.

The retrieval mechanism acts as an extensive database, supplying the generative model with a wealth of information. The generative model then utilises this information to formulate responses or content, further refined by the retrieved unstructured data, resulting in responses that are not only precise but also contextually enriched.

The Advantage of Pre-Trained LLMs

Constructing a Large Language Model (LLM) from the ground up is a daunting endeavour, often demanding considerable resources and time. RAG presents a powerful alternative, empowering users to exploit the potential of existing language models without undergoing the rigours of development from scratch.

For the Data Science team at Etops, this translates to the integration of cutting-edge AI functionalities within our suite of products and services. By leveraging established language models, we enable our clients to draw upon a vast repository of knowledge, ensuring that all generated content is not only accurate but contextually well-informed.

How DALL-E 3 imagines the future of AI & Wealth Management
The Evolution of AI Assistance with Etops

RAG represents a big leap forward in AI, offering sophisticated help with minimal training. It blends data retrieval with content creation to improve user experience. As data grows and clear communication becomes more crucial, an AI tool like RAG is a significant asset. It offers businesses and individuals a smart digital resource. Etops is dedicated to providing tools like Co-Pilot that strive to meet and support our clients' needs. We're improving Etops Co-Pilot every day and welcome your input to perfect it. Etops Co-Pilot will one day be as essential as a web browser and present a better way to understand, discover and browse your data.

This Blog was authored by Henry Valeyre, Robert Klöpsch & Markus Kessler.

Etops Santiago Schuppisser
Santiago
CPO - Etops