Recently, Citi Ventures made a move to secure its place in the future by investing in Anaconda, an artificial intelligence-based software company.
Citigroup, Citi Ventures’ parent company has utilized the popular open source software for several years across its enterprise. Although the investment size was not revealed, it was said to be important for three reasons including:
1.It acts as a sign that additional banks are integrating artificial intelligence
Large banking institutions including Ally Bank, BBVA, the Bank of America and Wells Fargo are some of the many large banks boasting AI deployments. According to the Market Structure Metrics Principal David B. Weiss, there is a continuously growing, five-year trend of banking institutions tactically deploying AI among other related technologies such as machine learning in a bid to target multiple processes in various parts of their businesses.
2.It serves as a feather in the cap of advocates for utilizing open source software for artificial intelligence
According to McWaters, a considerable part of this activity is carried out in an open source setting. He added that it shows that the algorithms and methodologies used to develop AI are somewhat commoditized.
3.It offers a window into Citi’s increasing AI use across the entire company
Ramneek Gupta, Citi Ventures’ co-head of venture investing, said that the company has been striving to create the technology stack, specifically at Citi in an effort of driving widespread adoption of machine learning within numerous functions and use cases.
According to Anaconda’s CEO Scott Collison, every leading banking institution in the United States including JPMorgan Chase, HSBC, Barclays among others utilize Anaconda ’s software. Citi is the recent addition to the list.
While some banking institutions use Anaconda’s software for stress-testing, others such as Citi utilize it for AML and credit evaluation. Even so, others use it for risk analysis, loan decisions, and treasury applications.
Its increased use and popularity among universities and students attracted Gupta’s team to Anaconda. According to him, when students who use Anaconda enter into the workforce, they would require a similar setting, and that increases the demand for the platform.
Gupta said that Anaconda’s enterprise license is easy to scale up. He said that carrying out tasks in production in a cloud like platform boasting thousands of nodes calls for a distinct version and scalable platform.
Impediments of Large-Scale AI
McWaters said that accessing large and at times exclusive data sets for training algorithms is the challenging bit of incorporating AI tools such as Anaconda in large organizations.
In a study conducted recently through a partnership between Deloitte and the World Economic Forum, banking institutions described siloed environments not only through raw data that required cleaning up but also using legacy systems that needed revamping for the cloud before embarking on machine learning deployment.
As such, McWater said that even though such types of capability acquisitions could prove useful, they have to be combined with some expensive, challenging and relatively low-key investments, specifically in getting the tech and data of these legacy entities up to speed before you can start agile methodologies deployment.