AI & data science: Open source makes NSE smart and secure

Published: December 10, 2018 3:58:29 AM

NSE’s shift to cloud-based AI platform enables it to clock a daily turnover of Rs 3,00,000 crore with 1.2 billion daily transactions

NSE’s move to adopt algorithmic trading and building it in-house using open source has reduced the impact cost on trading to 2 paise per transaction, whereas, the US and European markets operate at a cost of 4-5 paise per transaction.

By Srinath Srinivasan

The debate between open source and propreitary software has been going on ever since the idea of democratisation of computer software came into existence. Today, with the growing need for digitisation and automation, some institutions choose open source solutions to automate their systems.

National Stock Exchange of India (NSE which used V-SAT to transmit data securely in 1993, had shifted to Red Hat open source later. In the last few years it has been strengthening that partnership further with the integration of cloud infrastructure in its data systems to not simply improve data security, but also to implement Artificial Intelligence (AI) and data science in its systems. The implementation of cloud-based AI platform enables NSE to clock a daily turnover of Rs 3,00,000 crore with 1.2 billion daily transactions. It is the largest stock exchange in India in terms of market volume and market share.
Says Yatrik R Vin, CFO, NSE India, “There are certain cases on which we use open source’s capabilities extensively. They are risk management at client and investor level, cost reductions and making our systems talk to the public without manual intervention.” He reminisces that during the financial crisis of 2008, not a single rupee was affected, because of the risk management capabilities of the eight-sigma level open source core systems that were in use at NSE India.

To operate at scale with the public and at low costs, NSE India utilises cloud capabilities developed in collaboration with Red Hat open source. The solutions have the capability to deploy AI to monitor risks real time, as the transactions take place. The systems can detect anomalies which lead to a crash and thus prevent downtime. It also saves NSE from subsequent recovery procedures that can be detrimental to its operational costs and customer trust. NSE’s move to adopt algorithmic trading and building it in-house using open source has reduced the impact cost on trading to 2 paise per transaction, whereas, the US and European markets operate at a cost of 4-5 paise per transaction.

Dirk-Peter van Leeuwen, senior vice president and general manager, Red Hat, Asia Pacific region, believes that for data and money-intensive institutions like NSE, the security offered by open source is much stronger than that offered by propreitary solutions.

“If it has to be secure, it has to be open source. The security problems we encounter are mostly at propreitary solutions. It takes months and sometimes years for them to identify that they were attacked. Then, the recovery takes further time. And all the while, you have been exposed to these security risks,” says van Leeuwen. “In open source, the quality of source code is so much stronger. All eyes are on how can one test the codes, break them and strengthen them. If you share the way you do security with everyone around the world, you get the best security.”

Apart from adopting tech and making it democratic, both Vin and Leeuwen believe in this key value—Start small and don’t stop developing for successful projects.

Get live Stock Prices from BSE and NSE and latest NAV, portfolio of Mutual Funds, calculate your tax by Income Tax Calculator, know market’s Top Gainers, Top Losers & Best Equity Funds. Like us on Facebook and follow us on Twitter.

FinancialExpress_1x1_Imp_Desktop