Aerospike’s real-time data platform enables organisations to act instantly across billions of transactions while reducing server footprint by up to 80%. Its multi-cloud platform powers real-time applications with predictable sub-millisecond performance up to petabyte-scale with five-nines uptime with globally distributed, strongly consistent data.
“The Aerospike Real-time Data Platform is used by applications to process fast-changing data at high throughput (scale) to facilitate artificial intelligence and machine learning model-based decisions within a fixed SLA (typically in milliseconds or tens of milliseconds). Applications include fraud detection, recommendation engines, real-time bidding for ads, inter-bank money transfers, etc. Aerospike technology enables petabytes of data to be processed efficiently, thus improving the quality of AI/ML models generated as well as the quality of decisions made by applying those models in real-time,” said Aveekshith Bushan, Regional Director & General Manager, Asia Pacific, Aerospike.
Sign up for your weekly dose of what's up in emerging technology.
In an exclusive interview with Analytics India Magazine, Aveekshith spoke about ethical AI and how it’s embedded in their platform.
AIM: What explains the growing conversation around AI ethics, responsibility, and fairness?
Aveekshith Bushan: According to the Worldwide Artificial Intelligence Spending Guide from International Data Corporation (IDC) forecasts, global spending on AI systems will increase from USD 85.3 billion in 2021 to more than USD 204 billion in 2025. However, with the increase of AI in an organisation’s system to cope with the rising levels of online interactions, ethical biases and transparency issues arise. The need to develop an ethical charter that defines AI algorithms is not only a moral responsibility but also a business imperative.
While Aerospike does not ingest consumer data, our customers do while working on our real-time platform. So, they must implement and follow ethical standards to provide safety guidelines that can prevent risk for both business and human interactions.
AIM: Why is it the need of the hour?
Aveekshith Bushan: AI/ML and the data required to create desired outcomes sit at the heart of the digital transformation businesses are experiencing today. As more and more data become available, more and more AI-based applications will be created, and existing applications will have to handle extreme loads of data to make better decisions. It’s mission-critical for organisations to adhere to ethical standards and pace themselves accordingly to keep up with the increase in AI/ML processing. Government organisations, enterprises, and small-to-medium size businesses all rely on AI/ML computations to drive their “decisioning” in a manner compliant with diverse ethical, social, and legal norms. India’s Telecommunication Engineering Centre (TEC), the technical arm of the Department of Telecommunications (DoT), has started discussions to create a framework for a fair assessment of AI and ML systems to build public trust.
AIM: What’s the biggest challenge companies face while ensuring AI governance at scale?
Aveekshith Bushan: The biggest challenge is that an algorithm that is launched initially for small workloads needs to continue to work well as it scales. The millionth or billionth user must have the same experience as the first user. This means the platform needs to scale up and be elastic in its scale-out. Non-scalable platforms tend to stop a hyper-growth business in its tracks. For example, if fraud detection SLAs do not keep up with the growth in transactions, a significant number of transactions will have to proceed without a fraud score. This in turn will lead to more exposure to undetected fraud and could eventually result in huge losses. Using platforms like Aerospike which has a proven track record of delivering predictable performance at scale with high uptime would be a way to prevent disruptions to high growth businesses.
AIM: How do you protect consumer data?
Aveekshith Bushan: Digitisation comes at a cost. Each of us now leaves a trail of digital exhaust, an infinite stream of phone records, texts, browser histories, preferences, buying patterns, location, and other information that lives forever. Every time someone views a webpage, check-outs at the supermarket, receive an electricity meter report, gets a package that passes through a delivery checkpoint, scans an ID, or posts on social media –the digital trail grows.
And so does the need for greater security and privacy of data for consumer and other business use case data. There cannot be a trade-off between speed and scale and security and privacy, they must move forward as one. It is important that applications built to power these digital transactions provide a mechanism for security and privacy.
The Aerospike Real-time Data Platform provides a capability for cross datacenter replication (XDR). Aerospike’s XDR enables enterprises to create a global data hub (Figure attached) that automatically routes, and augments data captured anywhere in the data centre to wherever it’s needed – whether in Aerospike clusters or any other data repository. Typically, regulatory compliance is highly regional, and the new XDR provides the ability to manage regulatory requirements such as GDPR and CCPA on a regional basis.