Inside Citi Ventures’ AI Playbook: From Compliance Automation to Hyper-Personalization
As Citi’s venture capital arm, Citi Ventures scours a broad swath of startups from around the world to zero in on those that are generally aligned with the banking giant’s business strategy.
From that vantage point, Citi Ventures has a good view of the artificial intelligence (AI) advances percolating from the cutting edge of Silicon Valley and beyond.
In this exclusive interview, Jelena Zec, partner at Citi Ventures, shared the AI developments she is seeing not only from among the fund’s portfolio of 150 companies but also elsewhere. From regulatory compliance to ultra-personalized marketing offers, trends are starting to emerge.
Zec said she looks at three factors when choosing startups to invest in: the founder, a sufficiently large total addressable market (TAM) and the business model, in that order.
Citi Ventures invests in seed to Series C rounds, or early to late-stage businesses, with the average check ranging from $2 million to $15 million.
Zec cites the following emerging trends in AI and FinTech:
- Speeding up compliance procedures
A major trend emerging in financial services is the application of generative AI to regulatory compliance. Zec highlighted Norm AI, a company that trains specialized AI agents to assist SEC-regulated financial institutions.
“Every time an SEC-regulated entity — it can be Citi, it can be a hedge fund, it can be a private equity fund — publishes [content] externally, it has to be SEC compliant,” she said.
Norm AI streamlines the process by accelerating the review of marketing materials, flags content that is not compliant and recommends how to rephrase things to comply with current regulations.
This technology moves beyond simple rule-based systems to AI that can understand context and nuance in regulatory language. The technology accelerates the review process while enhancing accuracy and addresses a labor-intensive process in financial operations.
- Content generation with a brand’s voice
Another trend is AI-powered content generation specifically tailored to be faithful to a brand’s voice. Citi recently invested in Writer, which could be used by companies to communicate with clients with their own voice.
Writer can help craft emails telling clients about the latest market trends while adhering to company guidelines, for example, Zec said.
What distinguishes this trend from general AI content generation is its focus on compliance and brand consistency — critical concerns in highly regulated financial environments.
Read more: How Web3 Can Raise Customer Engagement in Rewards Programs
- Wealth adviser engagement with clients
The wealth management industry is seeing AI innovations as well. Zec pointed to startup Jump as an example of how generative AI is enhancing advisor efficiency by taking precise notes during meetings with clients.
Jump uses generative AI to make wealth advisors more efficient. “It starts by capturing notes during meetings with clients and gathering all of what is unstructured data to enrich the CRM system,” she said.
The technology allows advisors to focus entirely on client conversations while AI handles documentation, automatically uploads data to Salesforce or other CRM platforms and assists in drafting follow-up communications.
Notably, Jump is “building agentic workflows that will also be able to execute tedious tasks on behalf of the wealth advisor,” Zec said.
In the estate planning space, AI is solving the complex problem of digitizing non-standardized documents.
Zec cited Wealth.com, which uses gen AI to extract information from old estate trust documents that are old, in poor shape perhaps with coffee stains, or handwritten so they’re hard to read.
“They utilize gen AI to abstract information from old estate trust documents that are non-uniform and difficult to digitize,” Zec said.
- AI agent-to-agent payment systems
For some reason, there has been less innovation with payment systems until recently, Zec said. But she is seeing some trends emerging in this area.
One comes from a startup called Skyfire, which is building AI agents that can initiate payments using stablecoins as payment rails.
“You create an AI agent and empower that agent with access, setting certain parameters like transaction limits,” she said. These agents operate within defined parameters but can autonomously execute financial transactions.
The payment is made in stablecoins, cryptocurrencies pegged to a hard currency like the U.S. dollar, and the recipient converts it into hard currency. Zec said the use of stablecoins makes the transaction “super fast” and 24/7.
One concept from Skyfire that excites Zec is “KYA,” for know your agent — a twist from the traditional know your customer (KYC) and know your business (KYB) protocols in financial services.
Another company in this space, Payments, utilizes AI agents for executing mass payments. “They are utilizing AI agents … in big, massive payouts,” Zec said. An example is paying hundreds of gig workers all at once.
- Hyper-personalization
After years of discussion about personalizing in commerce, artificial intelligence is finally enabling true hyper-personalization in financial services. “With gen AI, that piece can finally fulfill its promise,” Zec said.
The convergence of vast customer data with real-time AI analysis capabilities is creating opportunities for personalization.
As an example, Zec pointed to OfferFit, which employs reinforcement learning to conduct what amounts to individualized A/B testing at scale — something previously impossible with traditional technologies.
“Think of it as A/B testing, but in real time, and you can do A/B testing on individuals, so every individual can get individual A/B testing at scale, thanks to technology,” she said.
See also: Retail Personalization Is About More Than Just Sending Offers
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
The post Inside Citi Ventures’ AI Playbook: From Compliance Automation to Hyper-Personalization appeared first on PYMNTS.com.