The biggest names in data management gathered in New York for A-Team Group’s 14th annual Data Management Summit New York City. In the second part of our account of the major discussion points at the day-long event, guest speakers and panellists talk self-service analytics, regulatory data and tooling implementation.
What are the challenges of delivering a successful self-service analytics?
Amin Zaman, senior vice president for client advisory and management at Curinos expressed little surprise in the poll responses to this question. The lack of accessibility to data and a lack of trust in data topped the audience answers in the poll that was set during the “Best practice approaches to self-service analytics to empower users and unlock data democratisation” panel discussion.
Zaman tempered his response by saying that firms that are starting with a “clean sheet” – when building data operations for new products or services, for instance – would be able to adopt self-service analytics easier because larger organisations with legacy technology structures would need to apply them to established lines of business.Another challenge is that self-service has evolved to mean different things to different parts of an organisation, said Adam Houhoulis, director, data management office at Federal Home Loan Bank of New York. That difference was starker between savvy and not-so tech savvy users, making it difficult to pitch self-service analytical products.
What aspects of data management for regulatory reporting are most challenging?
During the panel that considered best practice approaches to data management for regulatory reporting, this question received a plurality of responses, with data transformations and lineage marginally topping data sourcing across silos/legacy systems.
Moderator Dessa Glasser, independent board member at Oppenheimer, said that the list of challenges is long, sentiment echoed by ****Brett Utnick, head of regulatory operations at ****BMO Financial Group. Utnick added that he was unsurprised that respondents voted in large numbers for all possible answers.
Panellists agreed, however, that lineage was particularly important for satisfying regulators. Murali Duvapu of Scotiabank said the lender had automated processes that maintain the traceability of its regulatory data. Getting that piece right was critical because the company processes so many annual, quarterly and monthly reports, said the senior manager and data governance executive. It’s a “big challenge” Duvapu added, but insisted that Scotiabank had achieved its objective and is continually improving.
John Carroll, chief operating officer at Datactics, agreed that ensuring complete lineage is an important part of compliance regimes but cautioned that measuring the return on investment of its automation is difficult, making its implementation a difficult sell to boards.
The panel agreed too, that all responses offered for the next question – How should responsibilities for regulatory reporting be organisationally structured for optimal performance? – were equally valid. Nevertheless, a federated approach was preferred by a wide margin over a centralised or decentralised policy in the audience poll.
BMO’s ****Utnick said each approach would suit different businesses based on many factors but particularly its people; if a company’s data team can define its approach adequately, then it can make any strategy work.
What do you consider to be the biggest barriers to successful implementation of data technology and tooling?
The final poll question of the summit saw a close run between two responses: again, integration with legacy systems and data quality/trust. Cultural issues and user adoption also polled highly.
Moderator Julia Bardmesser, founder and chief executive of Data4Real, put the cat among the pigeons by suggesting that many of the barriers highlighted could be overcome with the implementation of a dashboard, which she argued would help embed tools into business processes.
Rimes head of AI product Theo Bell disagreed, suggesting tools integration with workflows would be a more efficient solution.
JPMorgan Chase has created a real-time dashboard from which the bank’s sales people are taking selling cues, said Linda Zhang, executive director, commercial and investment bank – data and analytics technology.
Bardmesser and Zhang also found common ground in the use of data observability software to monitor data pipelines. The moderator said this new class of tooling was particularly important in ensuring the integrity of regulatory data, while Zhang said that without such technology she wouldn’t be able to trace where pipeline breakdowns had occurred.
Subscribe to our newsletter
