
What happened
Circles, a Singapore-based technology company, used OpenAI's API and Codex to build an AI-native telco stack. In Singapore, this increased ARPU by 22% and reduced churn by 9%.
Why it matters
The results show AI can turn rich telco data into proactive, personalized customer experiences, moving beyond reactive support. Circles' AI Concierge, powered by CareX, now resolves 65% of customer service interactions autonomously.
What to watch
Circles is targeting 95% autonomous resolution as CareX expands into real-time voice support. The 22% ARPU increase and 9% churn reduction were measured in Singapore, so the test is whether these gains scale across 14 countries.
Summaries like this, in your inbox every morning.
Circles' approach addresses a common telco problem: data is rich but hard to act on in real time. By building an AI-native stack on OpenAI's API, the company turns usage, billing, and network signals into proactive actions like recommending roaming packs or preventing churn. Early results in Singapore are strong, with a 22% ARPU lift and 9% churn reduction, but these numbers come from comparing customers who received AI recommendations against those who did not, so the exact methodology and broader applicability are not detailed in the article.
The company is also using Codex internally to improve engineering speed, reporting a 29% increase in development efficiency. This dual use—customer-facing and internal—shows how OpenAI's tools can drive both revenue and productivity.
The key test is whether these gains can scale across Circles' 14-country footprint and extend to voice support. CareX's target of 95% autonomous resolution in voice is ambitious, and success will hinge on maintaining quality and cost control while expanding to more complex interactions. If Circles can replicate Singapore's results elsewhere, it could set a new standard for AI-powered telco experiences.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Vinci4D provides chip-design simulation software that trains on physics equations instead of customer data, an…
SEMI President and CEO Ajit Manocha told DIGITIMES that the semiconductor industry reached the US$1 trillion r…

CoreWeave fell more than 25% this week, its debt yield briefly rose above 16% on Thursday, and SK Hynix droppe…

Nvidia's AI Enterprise lists at $4,500 per GPU annually, and paid subscriptions add supported software, securi…

Anthropic told investors this week that it is making 85% gross margins, a figure the columnist calls a cash co…

Modal Labs, an AI infrastructure provider, raised $750 million in a Series D led by Accel at a $15.75 billion…
