The Singapore deep-tech company that taught machines to see a network fault coming - before the customer ever does.
Every large network hums with warning signs. Alarms flare, degrade, and clear thousands of times a day, and most of them mean nothing. Buried inside that noise is the one signal that matters: the fault that is about to take a cell tower, a backbone link, or an enterprise IoT deployment offline. Avanseus was built to find that signal.
Founded in Singapore in 2015, Avanseus Holdings set out with an unusually specific ambition - not to build another dashboard, but to predict network faults and performance degradations before they happen. Its flagship product, the Cognitive Assistant for Networks, forecasts failures 7 to 30 days in advance and recommends the likely root cause based on historical fault patterns.
The technical wager underneath the company was that you do not need years of labeled data to do this well. Avanseus developed a patented, neural-network-based unsupervised learning algorithm that can start predicting from as little as three to six months of historical alarms - and, in the process, tackled a stubborn problem in AI known as long temporal dependency.
That combination - fault prediction on minimal data, delivered non-intrusively across everything from legacy wireline to 5G - is what set Avanseus apart. In February 2026, Accenture acquired the company's AI technology to accelerate its telecom clients toward autonomous networks. For a roughly 46-person deep-tech team, it was a quiet, defensible kind of win.
"Accenture's acquisition of our AI solution marks an important next chapter for the technology we have built."
Bhargab Mitra - Co-founder & CEO, Avanseus
The Cognitive Assistant for Networks watches historical fault data, learns each network's own patterns, and turns that into an advance warning - plus a ranked view of which faults actually matter.
Configurable 7-30 day advance warning, built from as little as 3-6 months of historical alarm data.
Cloud-native AI/ML platform that predicts faults and degradations 7-30 days ahead and recommends root causes. Non-intrusive, spanning mobile, fixed, IP/MPLS backbone, transmission and enterprise IoT.
A patented neural-network-based unsupervised learning engine that forecasts from as little as 3-6 months of alarms and addresses long temporal dependency in AI.
Models for anomaly detection, cross-domain correlation, impact analysis and optimization to lift operational efficiency in complex networks.
Fault impact estimation, impact scoring and prioritization so operations teams act on the faults that matter most - not the ones that shout loudest.
Where it fits: Avanseus sits in the AIOps and network-analytics market alongside players such as Anodot, Moogsoft, BigPanda and Nokia's AI tooling - but with a sharper focus on advance fault prediction as a building block for autonomous networks. That focus is precisely what drew Accenture, which folded the technology into its cognitive network platform.
Bhargab Mitra and co-founders launch Avanseus, backed by a US$2.5M seed round from SEEDS Capital.
The universal prediction algorithm - unsupervised, neural-network-based - takes shape.
Convertible-note financing led by TNB Aura, with SEEDS Capital participating, as Cognitive Assistant for Networks gains traction.
A roughly US$2.4M Series A with TNB Aura and SEEDS Capital to scale the platform.
Listed on Red Hat and VMware marketplaces and certified on Red Hat OpenShift.
Accenture acquires the AI technology to accelerate autonomous-network journeys for telecom clients.
Avanseus builds AI-based predictive maintenance software that forecasts faults and performance degradations in telecom and industrial IoT networks, and recommends likely root causes, so operators can fix issues before they cause outages.
Its flagship product is the Cognitive Assistant for Networks (CAN), a cloud-native AI/ML platform that predicts network faults 7 to 30 days in advance using a patented unsupervised learning algorithm.
Avanseus was founded in 2015 and is headquartered in Singapore, with an R&D center in Bengaluru, India. Bhargab Mitra is co-founder and CEO; other co-founders include Giuseppe Donagemma (Chairman), Chiranjib Bhandary, Mei Lan Ng and Rajendra Panda.
Yes. In February 2026, Accenture announced it acquired Avanseus's advanced AI technology to strengthen its cognitive network platform and help communications companies accelerate their journeys toward autonomous networks.
Avanseus raised roughly US$6.2M in total, including a US$2.5M seed, a US$1.3M convertible-note bridge in 2019 led by TNB Aura with SEEDS Capital, and a Series A of around US$2.4M.