Two financial technology companies are taking different approaches to embedding AI into capital markets infrastructure — one through a client-facing research assistant, the other by rewiring the development layer underneath. 

Tradeweb and 3forge, which is partly owned by Morgan Stanley, are two fintechs that work with nearly all major FIs in the United States and are focused on developing AI tools to help banks increase efficiency in capital markets operations, executives told FinAi News.

When asked about record market revenue of $7 billion, up 17% year over year in the second quarter, Citi’s Chief Executive Jane Fraser said during the earnings call on July 14 that the bank is ” deliberately investing for long-term growth and for improved returns.” 

Wells Fargo’s CEO Charlie Schraf echoed that sentiment on July 14 during the bank’s Q2 earnings call, stating that the investment the bank is making in its Wealth and Investment Management group is intended to “improve operating discipline” and grow business organically. 

As banks increase focus on capital markets operations, fintech like Tradeweb and 3forge are developing AI features and tools to aid the increased focus.

New York-based Tradeweb, an electronic bond trading platform, last month launched Tara (Trading AI Research Assistant), Izzy Conlin, head of strategy and solutions for global markets, told FinAi News, adding that Tara is “the first iteration [of tools] in our AI journey.” 

The tool is designed to give buy-side traders and dealers instant access to data points, Conlin said. In bond markets, where trading remains largely over the counter and pricing can be opaque, even basic queries — such as which dealers offered the best execution costs over the past week — can require significant legwork. 

“You’d think it’s an easy thing to do,” Conlin said, “but because data is so fragmented in this opaque fixed income world, it’s very difficult.” 

The tech 

Tara is built on Amazon Bedrock and uses open-source Python platform LangChain for memory, Conlin said.  

It covers U.S. credit markets, with European markets in beta testing. Tradeweb plans to extend the tool to U.S. Treasuries, global swaps and ETFs by yearend, Conlin said. 

Large asset managers and hedge funds are among the early adopters, though Conlin declined to name clients.  

She said many Tradeweb clients are interested in using the tool and the adoption curve has been steep since launch. 

FIs that use Tradeweb’s solutions include: 

• Goldman Sachs;  

• JPMorgan;  

• Morgan Stanley; and  

• Barclays. 

Tradeweb’s edge lies in more than two decades of proprietary fixed income data, including records of trades that were initiated but never completed, she said. 

3forge’s infrastructure layer 

Where Tradeweb is building toward the trader, 3forge is building underneath existing infrastructure. 

The New York-based data analysis company, founded in 2012, has integrated AI into its platform to allow clients to quickly build real-time dashboards without writing extensive code, founder and CEO Robert Cooke told FinAi News. 

The more consequential AI opportunity is not in spinning up new applications, but in making existing systems more resilient, he said.  

“Things that were very complex… it was just too hard to solidify,” Cooke said. “Suddenly that becomes attainable through AI.” 

The company’s AI-assisted dashboard builder, in beta with tier-one bank clients ahead of a general release this summer, allows users to submit a screenshot of a desired interface and receive a working, real-time application in return, he said. 

What previously carried a multiweek learning curve can now be completed in under a day, Cooke said, representing a five- to tenfold productivity improvement. 

FIs that use 3forge’s solutions include Morgan Stanley and Citi.