Converting raw data into fresh insights

This article first appeared in the National Research Foundation Magazine, Issue 8 – January 2020.

The management team of Niometrics

The management team of Niometrics comprising Founder and Chief Executive Officer Kostas Anagnostakis (centre), Chief Technology Officer Periklis Akritidis (left) and Chief Operating Officer Zach Tigkas.

A*ccelerate is the research commercialisation arm of the Agency for Science, Technology and Research (A*STAR). It has supported deeptech startup Niometrics, one of the fastest growing tech company in Singapore. We chat with Dr Kostas Anagnostakis, scientist-turned-founder of the company.

Niometrics wants to help telcos monetise the data that they collect through serving their customers. How does Niometrics do that?

Telcos have witnessed significant value that can be created and captured by them through the data they acquire. That has prompted them to pursue new sources of competitive advantage within the technology space they participate in.

To reposition and reclaim relevance within the broader digital ecosystem, telcos are seeking a new class of data enablement, one that can uplift them from passive observers into more active agents of the experiences they help to fulfil.

Niometrics’ network analytics technology helps them to do just that. It produces data intelligence that lets telcos drive customer experience improvements, foster product innovation, fuel digital partnerships and establish innovative data platforms.

It does so by detecting, analysing and deriving insights from the huge volumes of data traffic that flows through telco networks. Niometrics’ network analytics technology transforms all that data into higher-order insights that can feed innovative initiatives for telcos to pursue.

What do you think sets Niometrics apart from other data analytics companies?

Niometrics’ network analytics technology features a unique full-stack software approach, with end-to-end control over how different components come together to deliver deeper and faster analytical performance. Unlike traditional solutions, Niometrics focuses on vertical specialisation for network analytics. We build a seamless value chain of data extraction, interaction, mediation and visualisation, avoiding the integration pitfalls that companies face with fragmented, multivendor tools.

In addition, our system uses a hardware more efficiently, and this results in lower footprint requirement. This is because our network analytics platform delivers a big data experience out of significantly smaller server requirements, demanding up to 80% less hardware than other alternatives would. In an environment where hardware investments can reach millions of dollars, while weakening insights is not an acceptable alternative to minimising costs, that degree of efficiency makes a big difference.

As a very important bonus, that smaller hardware requirement also translates into significantly lower energy consumption, doing its part in helping information technology platforms to become less energy-hungry and greener.

Lastly, we provide real-time analytical data. Our analytics give communication service providers the advantage of speed – going from trillions of raw metadata to first-order events in a split second – without compromising on accessibility, retainability, and timeliness of critical data. It is highly performance-scalable so that infrastructure costs do not commensurate with data traffic growth.


Niometrics’ network analytics technology transforms all that data into higher-order insights that can feed innovative initiatives for telcos to pursue.


Dr Kostas Anagnostakis on Niometrics’ key technology use.


You developed Niometrics’ network traffic analysis engine when working as a research scientist at A*STAR. What was the process like with starting and growing a research spin-off?

Moving from research to spin-off mode injected a whole new world of practical considerations into me and my co-founders’ day-to-day operations. From a more acute drive to find applications to our research to the engineering of a sustainable business model around it, we paid attention to every little reality check.

A*ccelerate made valuable contributions at various stages of our development. For example, it proposed for Niometrics to apply its signature validation technology to the then up-and-coming applications in deep packet inspection (DPI), and brought in industry collaborators to advise on potential commercial models around them.

In addition, A*ccelerate facilitated the team’s pitching to potential pilot users, which resulted in one of the first large-scale deployments for the company. The agency also gap-funded the development of Niometrics’ algorithm into reusable software engines/ SDKs.

It’s been 10 years since Niometrics was spun off using A*STAR’s network traffic analytics software. What is the end goal for Niometric, and when do you hope to achieve it?

Our end goal is to maximise the impact that our technology can have in making deeper sense of digital moments. On that, we believe there is still a lot to be achieved. These include expanding the breadth of intelligence that we generate for telcos and applying our core technology into adjacent areas, like cybersecurity. We plan to do these through investments or other corporate milestones. We have been lucky enough to be profitable since our inception. This allows us to chase our plans in a very organic manner.

What, in your opinion, is the future of network analytics?

First, there is a need to build privacy protection through privacy engineering. A network analytics platform that does not follow privacy engineering methods risks overstepping the privacy standards of both today and tomorrow. From basic anonymisation to more structural privacy-protection mechanisms, network analytics is expected to reconcile its ability to generate intelligence-driven impact while delivering safe data experiences.

Second, we need to enhance 5G deployment with pointed intelligence guidance. 5G is increasing the complexity involved in operating and managing mobile communication networks. By enabling ubiquitous IoT adoption and creating a standard framework for ultra-low latency applications through edge computing, a 5G network reality will require a great deal of optimisation, largely driven by network analytics.

Niometrics has expanded its footprint internationally with offices in countries such as Singapore, Indonesia, Australia, Greece, and Romania. What has been your most surprising finding about how data is being utilised by companies in different countries?

Perhaps the most surprising finding is that in all countries, no matter where they are in their digital maturity journeys, combining intelligence generation with privacy protection has become a non-negotiable imperative. Those two forces – the need to advance on data insights without compromising on privacy – are becoming more and more indissociable.

And that’s how it should be. In Niometrics, we believe that progress in data analytics must come at no cost to privacy. And that advances in privacy protection must not exact any toll on the richness of data analytics. Both must drive each other forward in mutually reinforcing fashion.

What valuable lesson did you gain as a researcher turned-entrepreneur?

That becoming a researcher-turned-entrepreneur means, first and foremost, surrounding yourself early on with the right people.

Two senior members of our pioneering team (our current CTO, Periklis Akritidis, and our current COO, Zach Tigkas) joined Niometrics in a very early stage, making instrumental contributions to our technology development and business expansion.

While researchers may be used to working predominantly alone, entrepreneurs must tirelessly seek out the support and collaboration of others. It is only through collective efforts, combining different skills and personalities, that ideas can genuinely turn into impact.