By Prashanth Kaddi (Partner); Soumya Hansa (Associate Director) and Raghu A, (Associate Director), Deloitte India
Analytics has arguably become one the biggest catalysts of change not only in technology but business interactions and business models. This trend has been further accelerated by the ongoing need for virtual experiences in the times today since the pandemic. The challenge has been in generating and sustaining insights due to several issues – availability and quality of data, need for a number of “smart” data scientists to generate insights and also the need for an ongoing specialist team to supply insights into the organization. Hence, relatively it has been challenging to democratize data analytics and adopt in the “non-data oriented” workforce of the enterprise.
While it may look like a circular solution, the answer to this seems to be emanating from analytics itself – Augmented Analytics.This field helps explore and analyse data and manage the cycle of data to insights to action in a more automated way, aiming to democratize the data analytics process for data, data science as well as business users.
The case for Augmented Analytics
Data adds value to its potential in the enterprise only when converted into actionable insights. To convert data to actionable insights, the following are the typical steps:
• Digitization of data for analysis – Data collection / preparation from multiple data sources(reliable governed dataset)
• Digitization of insights – Descriptive analysis and definition of relevant KPIs
• Digitization of cognition / prediction – Development of machine learning modelsand synthesis of information for generation of predictive insights
• Digitization of dissemination – Communication of insights meaningfully to stakeholders with action orientation
These steps require a variety of skillsets, which are often scarce, expensive to hire and also difficult to retain. Therefore, irrespective of the level of ambition to leverage data analytics to make decisions, many organizations have succeeded in varying degrees of maturity and adoption.
This is where Augmented Analytics comes in – it addresses this imbalance by leveraging Artificial Intelligence (AI) and Machine Learning (ML) to assist with the data-to-insights-to-action process.
Augmented Analytics to accelerate the Data-Insights-Action journey
For organizations to embed analytics into their organizations, augmented analytics can be an easy win that solves some of the pressing problems already faced by data, reporting and analytics teams:
• Digitization of data for analysis –Intelligently accelerate data quality, profiling, wrangling and enrichment of data, more so in developing markets where data is often not completely reliable in organizations. Use of both analytics and technologies such as graph databases makes the process of readying data for analysis proactive and predictive rather than playing “defense”.
• Digitization of insights –Data discovery via visualization, including Natural Language Query (NLQ) and Natural Language Generated (NLG) techniques towards interpretation of insights in the specific context. Forecasts and contextualized insights through use of predictive algorithms can significantly reduce barriers to value that may typically exist.
• Digitization of cognition / prediction–Data scientists are becoming the hardest tribe to recruit, keep productive and retain in the data analytics value chain. Leveraging AI / ML techniques to this process to build models such as feature engineering, model selection, model execution and interpretation. Continuous improvement and development of models post deployment can be achieved using ML-Ops techniques.
• Digitization of dissemination– The experience in most organizations is that it is usually difficult to get senior leaders to consume insights by logging into or getting mailers of reports.Leapfrogging, conversational AI can deliver a personal digital assistant type experience, through a combination of Natural
Language based techniques, where senior management is looking for actionable insights or prescriptive insights without the need to look at reports or dashboards.
Business ROI for such initiatives typically emanates from both the value lever of better decisions as also the efficiency lever from reducing the effort and cost to deliver analytics. In conjunction, it is critical for enterprises to invest in data analytics training at senior/mid management to drive adoption and ensure champions for insight-driven decisions.
Augmented Analytics can help drive towards the ultimate goal of significantly increasing usage of data and insights in decision making across the enterprise, through an intuitive and action oriented experience, while reducing the cost of generation of insights.