Foundational models such as BERT and GPT-n series have changed the world overnight – at least in perception. But not quite for the enterprises, for whom finding the right technology partner is more important than the technology itself. Today, customers no longer consider just one application on a single technology, they now evaluate multiple technologies and integrate them to facilitate expansion.
There may be multiple APIs currently available for integrating applications, but the use case that you’re trying to solve depends on the domain expertise of the provider. Conducting a Proof of Concept (POC) and testing the technologies is crucial to ensure that they are effective for decision-making and beneficial for the business. Otherwise, there are many API integrations available in the market.
Domain expertise is important to build a complete ecosystem that can scale. This can help businesses leverage relevant knowledge and datasets to develop custom solutions. This is why enterprises look for enablers that can bring in the domain expertise for particular use cases.
Data challenges everywhere
Today, ESRI, a geographic information system (GIS) software company, for instance, has data from all over India pertaining to, say, what is being cultivated where. Deepak Kolekar, AVP & head of IT at Godrej Agrovet, explains that commodities such as maize or soybean extraction make up 60% of their raw material cost. If they can leverage ESRI’s data to source material from specific locations and bring down their costs, it will ultimately benefit consumers and give them a competitive advantage. This is just one example of multiple possibilities that can be explored.
Godrej’s data journey began almost 13 to 14 years ago, and, Kolekar maintains, now is the time for them to leverage the data that has been accumulated over the years. They are exploring the implementation of AI/ML technologies for exceptional reporting, rather than for daily routine use. In addition to this, they have developed planning solutions and are seeking to enhance it with image-based processing for their oil palm plantation business. Moreover, they have conducted a pilot project to predict which customers are likely to default on payments, which can aid in better decision-making.
The pilot was primarily launched in the B2B segment of their business, where they were able to achieve up to a 70% prediction rate and provide timely alerts to both the finance and operations teams. In order to validate the accuracy of the model, they instructed the finance teams to simulate the predictions using traditional methods in Excel. This exercise gave them confidence that such technologies are certainly capable of providing valuable assistance.
On the other hand, for a company like boAt, the journey with data is still in its nascent stages. “We have limited historical data, which means we are not yet able to utilise predictive or prescriptive insights. Our primary use case would have been demand forecasting, but given that 80-85% of our sales come from marketplaces like Amazon and Flipkart which are extremely fluctuating, demand sensing is not effective in that context,” explained Shashwat Singh, CIO at boAt.
At present, the company is concentrating on big data challenges such as identifying customer needs to aid in the New Product Introduction (NPI) process. This task involves analysing a significant amount of text and does not necessitate an extensive data history.
Role of tech enablers
One of the challenges that companies encounter today is how to utilise data effectively as per their business needs. According to a global survey conducted by Oracle and Seth Stephens-Davidowitz, 91% of respondents in India reported a ten-fold increase in the number of decisions they make every day over the past three years. As individuals attempt to navigate this increased decision-making, 90% reported being inundated with more data from various sources than ever before.
“Some interesting findings we came across was that respondents who wanted technological assistance also said that the technology should know its workflow and what it is trying to accomplish,” Joey Fitts, vice president, Analytics Product Strategy, Oracle told ET.
The current major players in the ERP solutions market are SAP, Microsoft Dynamics, and Oracle Fusion. Regarding boAt’s decision to choose SAP, Singh cited three reasons. Firstly, the feature functionality gap between SAP and the other competitors they evaluated was the smallest. Secondly, the strong partner ecosystem with SAP was a key driver for them. Thirdly, the RISE program by SAP was an attractive feature to them, as it enables customers to focus on their core business while outsourcing the overheads associated with managing the infrastructure and uptime.