Snowflake’s ASEAN MD Satchit Joglekar stated, “What we are seeing is, only 23-percent of ASEAN businesses in general, are truly transforming with AI. The reason is because a lot of initiatives that are happening are very surface level, for example a chatbot (deployment). Some do say, yes, there is business impact, but… is it truly transformative to the business?
“The answer is no.”
Classic gaps
There are many significant infrastructure and data foundation gaps that need to be addressed if we truly want AI to be truly transformative.
The ASEAN MD opined that classic enterprise problems like the lack of quality data infrastructure, fragile data governance, and non-seamless integration with existing legacy systems, is currently limiting AI to only tactical applications.
What is the role an AI data cloud has to have for visibility into whether AI is actually delivering value?
Satchit elaborated with a hypothetical customer question that goes like this, “My data is currently in different places, in different applications and silos… HR data is in Workday… ITSM data is in ServiceNow… supply chain information is in SAP… CRM is in Salesforce… so, what are you doing as the enterprise AI Data Cloud to help me unlock this value and bring all of this together?”
He pointed out that an AI data cloud provider like Snowflake is actually at the convergence point of different enterprise systems because it partners with the abovementioned technology brands across an enterprise stack to unlock semantically rich data.
“We are partnering with these organisations to be able to very easily, and within seconds, make those data sets from each of all those applications appear natively within Snowflake, and then join it with other enterprise data.”
Delighting customers with data-driven products
Globe Telecom is one example of a customer that cleaned, governed, secured, and unified customer data in one place, the Snowflake AI Data Cloud, and then ‘activated’ the data to serve their customers better. Migrating the data from legacy systems to Snowflake, was done within 53 days, no mean feat.
“In the past, Globe’s marketing team was not able to drive true personalisation because of different data silos… now they are able to personalise down to even an individual level to recommend Next Best offers, and customise telco plans,” according to Satchit.
Of interest to this EITN editor, is Snowflake’s public marketplace with thousands of economic, healthcare, demographic, and even financial markets datasets. “These are already cleansed, business-context ready, and able to join with your enterprise data within seconds.”
We’re investing in the community across ASEAN – from students to developers to executives – to upskill a new generation on data and AI, not just in one country, but across all our major markets.
Satchit Joglekar
Grab, the e-hailing provider, is one example of a client which shares its data product, Grab Maps, on Snowflake’s public marketplace. According to Satchit, this product is able to detect and map out even alleyways between buildings, a useful feature that Google Maps does not have.
Community and the power of network
Internal marketplaces, a private, enterprise grade catalogue of data products, models, and analytic tools shared within a single organisation or conglomerate group, are helpful in turning siloed data into governed capabilities that are repeatable and optimised across the whole group of companies. For example, unified personalisation, loyalty programmes, and cross-sell offers.
Satchit observed that these marketplaces also enable secure data sharing and value aggregation to be able to serve customers in a better and holistic way.
Upskilling and enablement at all levels are also absolutely critical to be able to take full advantage of AI technologies. “We’re investing in the community across ASEAN – from students to developers to executives – to upskill a new generation on data and AI, not just in one country, but across all our major markets.”
Data matters
Satchit observed organisations that are truly transforming with AI have embedded it into core processes like personalisation, risk management, supply chain, and therefore have potential to generate recurring business value.
The technology has to go beyond AI model metrics like accuracy and latency. It has to enable business-level measurements, and better yet, tie AI activity to business outcomes.
Business leaders like Satchit are also starting to come forth with the idea that the data matters.
“Firstly, take a data-first, and not an AI-first approach, right? Let’s get back to basics… prioritise the data foundations before your AI implementation. And in general, let’s all recognise that there is no AI strategy without a data strategy.”
