CBA backs no-code platform to expand use of AI – Finance – Strategy – Cloud – Software

CBA sees its investment in no-code machine learning platform maker as a milestone that will allow anyone in the group to set up an “incredibly predictive model literally at the push of a button.” “.

In this week’s episode of iT news podcastCBA’s Chief Decision Scientist Dan Jermyn discusses the bank’s efforts and mindset in “democratizing” AI, as well as some of the key wins in reaching out to the clients that its specific field – decision science – has won to date.

Democratization is a key driver of ABC’s support for

The bank is placing a sizable, albeit measured, bet on’s technology: to become its “exclusive financial services partner in Australia and New Zealand” and make the provider’s “AI Cloud” available across the organization .’s tooling is part of a larger initiative that sees CBA building a “centralized AI capability that caters to everyone.”

“We’re an AI-driven company these days, and to do that, we need to empower many, many more of our people with access to great intelligence and insights so we can better serve customers. “, Jermyn says.

Jermyn compares to an “AI for AI’s sake”.

“If you think about how data science is traditionally done in most places right now, you take a problem area, you get the data surrounding that problem area, and you take a data scientist, and they’re trying to build predictive models,” he said.

“The data scientist cleans the data, thinks about different kinds of techniques, codes the model himself, gets a result, and then trains the model [further]. They then tune the hyperparameters [variables used in algorithmic training]and they can go back and try another technique.

“This process involves a real expert, but it can take several weeks, even months sometimes, to keep refining continuously.

“What does is automate this process, so it will look at every possible combination of the latest and greatest in modeling and data science techniques, iterate through them all automatically, tune features on the fly, and will combine several different patterns to get the right result.

“And what you find there is that you can get an incredibly predictive model literally at the push of a button – you provide the data, you mark the outcome you want to try to predict, and then the technology will automatically launch an incredibly predictive data model so that we can then use it to be relevant to our customers.”

With an organization the size of CBA, the ability to automatically generate a model best suited to the analysis task at hand is a powerful concept in itself.

It also offers CDA the opportunity to quickly expand the potential benefits of data science, as there is a much lower barrier to entry and not all use cases require a data science specialist. data science.

“ was such an attractive partner for us because it democratizes and lowers the barrier to entry for analysts or modelers who don’t necessarily have a data science background to produce more meaningful insights for themselves. and their business partners than they would otherwise be in the business,” Jermyn said.

“In the context of… the size of the CBA organization, you can see how this enables us to deliver much better business and customer outcomes by enabling people to have access to tools that previously would have been beyond of their skills or abilities.

“Every time we use our best people [on] our most difficult and challenging areas to produce better results, we achieve great results. Now we are able to do this on a scale that far exceeds what we have been able to do in the past.

“It opens up new opportunities for things we’ve never done before, and it’s really exciting for our customers.”

Although this may lead to a proliferation of models, one of the attractions of the platform is that it can be configured with railings that “govern and control the application of this technology”, and the complete end-to-end process of how a model was created can be fully audited.

“It comes with standard enterprise checks around explainability, so how do we understand how models produce the results they produce? ; bias detection, so how can we ensure we don’t have disparity in how we serve different subsets of customers? ; and really robust controls for anything that goes live so we know what’s been released, when and under what circumstances,” he said.

“We can audit it, and that way we can make sure we have a really strong framework for how we serve customers, making sure it’s consistent with our goal of building a better future for all.”

Jermyn adds that this is not materially different from the scrutiny already applied to bank modeling.

“All models we use within the company are now subject to our Group Model Policy,” Jermyn said.

“The bank has been using models for decades – all banks have – and so we already have a very robust framework in place that handles the documentation of the models: where are they located? Where are they going to live? data features do they use, how often are they audited and verified, etc.

“There is already an incredibly robust framework. Our AI models are no different: they go through the exact same robust governance, and we also check out some of the cool new things happening with AI around explainability and disparity. »

However, the addition of should simplify this process by automatically documenting model compliance, making it easier to audit models from the start.

Access to top talent

The partnership with also gives CBA access to elite data science talent, in the form of Kaggle Grandmasters: the highest-ranking people in the online community owned by Google and the platform. form of competition for data scientists. points out in its marketing that “the top 20 Kaggle grandmasters in the world (the community of top machine learning practitioners and data scientists) are creators” of the company.

Access to these people was a calling card for the ABC; the bank said late last year that it wanted these “machine learning engineers and product specialists to work full-time on developing new AI solutions.”

“Suddenly we literally have access to the best data scientists around the world working on problems for CBA and for our customers,” Jermyn said.

“It’s an incredibly exciting thing for us to be a part of.”

Access to’s resources will augment CBA’s considerable data and analytics teams and capabilities.

Currently, CBA has “a few thousand” people in its data and analytics function, which encompasses “a few hundred” people in Jermyn’s own subfield of decision science.

The decision science team includes data scientists, AI professionals, experimental scientists, and behavioral scientists, among other roles.

“The idea behind decision science was to bring together a mix of people with a wide range of experiences and ideas…to achieve a better outcome than would otherwise be the case,” Jermyn said.

Decision Science has already produced several flagship customer-facing services, including the Customer Engagement Engine (CEE) that predicts the next best conversation bankers should have with a customer, and the Benefits Finder, that encourages bank customers to request refunds or discounts. are entitled to but may otherwise be unaware of them.

Subscribe to iTnews podcast on Apple Podcasts, Google Podcasts, Spotify, Amazon Podcasts or wherever you find good podcasts. New episodes will be released every Monday.

Previous A Morinville student wins gold in the national trades competition
Next Council partners with leading homeless charity to deliver learning and life skills