TD Securities Taps Layer 6 and Openai to provide real -time real -time information to sales and trading teams

Build a td ai

Bosman said the idea of AI assistant came from a member of the stock sales team. Fortunately, the bank has a platform called TD Invent, where employees can carry ideas, and the innovative leadership team can evaluate the projects responsibly.

« Someone in our stock sales desk and quite said I had this idea and brought it to TD Invent, » Bosman said. « What I loved most about it is when you build something super magical, you don’t have to go out and sell or put a face on it. People come in and tell us, » We want it, we need it or we have ideas, « and it’s really the best when we are able to combine our investments in data, cloud and infrastructure. »

TD Security has built TD AI’s virtual assistant using Openai GPT models. Bosman said TD works with its technology teams and the Canadian company Ai Layer 6, which the bank acquired in 2018, as well as with other strategic partnerships. The assistant integrates with the bank’s cloud infrastructure, which allows her to access internal research documents and market data, such as 13F documents and historical equity data.

Bosman calls a TDS AI system for managing knowledge, a term that usually covers its ability to retrieve by extracting increased generation processes (RAG), summarizes and synthesizes information in « short context generalizations and insights », so its sales teams can answer customer questions.

TD AI’s virtual assistant also gives users with access to the TD Bank Foundation model, TD AI Prism.

The model, launched in June, is used throughout the bank, not just for TD securities. During the start, the bank said TD AI Prism would improve the forecast efficiency of TD Bank applications by processing 100 times more data, replacing models with one architecture and guaranteeing internal customer data.

« Development has created unique challenges as the Gen AI was relatively new to the organization when the initiative began requiring careful management and control navigation, » Bosman said. « However, the project has successfully collected a variety of teams throughout the enterprise, encouraging cooperation to provide an avant -garde decision. »

He added that one of the hallmarks is his ability to text to SQL, which converts natural language promotions into SQL requests.

In order to train the assistant, Bosman said TD securities had developed optimization to facilitate the process.

« With optimizations looking for patents in the rapid recording of engineering and dynamic several shots, we have successfully achieved the desired business performance through contextual training, » Bosman said. « As a result, it was not required to refine the basic OPENAI model for interaction with both unstructured and table sets of data. »

Banks slowly enter the agency

TD Bank and TD securities, of course, are not the only banks interested in expanding from assistants to AI agents.

Bny told Venturebeat that it began to offer multi -agricultural solutions to its commercial teams to help answer customer questions, such as those related to foreign currency maintenance. Wells Fargo also noted an increase in the use of its internal AI assistant. For its car sales customers, Capital One has built an agent that helps them sell more cars.

Many of these cases have occurred after months of pilot testing, as in any other industry; However, financial institutions carry the additional weight of the strict confidentiality of customer data and trust responsibilities.

TD Securities’ Bosman noted that many employees, even by the bank’s business, are increasingly familiar with instruments like Chatgpt. The challenge with assistants and agents of pilot tests lies less in their training for tools, but in establishing the best practices for using assistants, integrating them into existing work processes, understanding their restrictions and how people can provide feedback to mitigate hallucinations.

In the end, Bosman said the assistant would become something, even his users outside the bank would like to use TD securities when interacting.

« My vision is that we see AI as something that can add value to us, but also for external customers in the bank. At the moment, this is a huge opportunity for us to manage a stronger customer experience and to provide a better experience in the colleague, » Bosman said.

Daily information on business use cases with VB Daily

If you want to impress your boss, VB Daily covers you. We give you the inner spoon for what companies do with generative AI, from regulatory changes to practical implementation, so you can share information about the maximum return on investment.

Read our Privacy Policy

Thanks for subscribing. See more VB newsletters here.

An error occurred.