Latest news with #FrontierTechHub

Business Standard

Data errors should be minimal, need to follow Japanese standards: Niti CEO

BVR Subrahmanyam says India must drive down error rates in welfare delivery to Japanese precision levels to avoid exclusions, fund loss and policy distortion Dhruvaksh Saha New Delhi Niti Aayog chief executive officer BVR Subrahmanyam on Tuesday said data errors cost vulnerable citizens significantly, and India must move beyond considering above-par quality as good enough—striving instead for complete accuracy along Japanese standards. Citing the example of LPG cylinder beneficiaries, he said: 'For 10 crore [100 million] people who have gas cylinders in India, even if the error is five per cent, that's 50 lakh [5 million] people. So, 50 lakh accounts have a bounce-back, and there is no transfer due to some or the other errors. This is noise [in the system]. So, in India, we can't go by percentages. We'll have to go to the Japanese quality level of 0.0001 per cent error rate. That's the level you need to drive things down to, and that's where I think data quality becomes important.' The CEO was speaking at the launch of the third issue of the quarterly tech insights report by Niti Aayog's Frontier Tech Hub, which called for improved data accuracy to enhance the quality and outcomes of government support programmes. 'Across ministries and states, platforms are built as separate silos that store data in clashing formats and use incompatible identifiers, forcing even routine joins to be stitched together by hand. Many of these systems run on ageing back-ends that lack basic validation rules, audit trails or version control, so a small tweak in one module can topple an entire workflow,' the report said. Under the data quality monitoring framework proposed in the report, it advocates self-assessment of current datasets and the formulation of a scorecard for data quality and maturity. 'When a single erroneous digit can delay a benefit, or a duplicate record inflates welfare outlays, the true cost of poor data becomes painfully apparent—impacting budgets, distorting policy and eroding the faith citizens place in our digital systems. The move from 'scale to precision' is not just an aspiration; it is a national imperative,' said Saurabh Garg, secretary, Ministry of Statistics and Programme Implementation (MoSPI). Subrahmanyam added that an overhaul of current data systems is needed if substantial gains are to be realised. 'The big gains come when we recraft and restructure the old systems. And that's when we get the great leap forward,' he said.