What is MCP and why does it matter for AI? Model Context Protocol (MCP), open-sourced by Anthropic last November, is a tool that standardizes how AI models tap into scattered data sources. It acts as a kind of universal port that lets AI agents grab the data they need when they need it without custom integrations. It lets agents work autonomously, share context, and communicate with each other—and underpins the coming wave of AI-driven services. Has your team given it a try?
Everyone’s talking about it, but how do you apply agentic AI to telecoms? These autonomous systems promise to push telecom automation beyond today’s predictive or GenAI tools, tackling tasks from real-time RAN optimization and fraud detection to customer care. Early proofs of concept, like Google Cloud-Deutsche Telekom’s AI-driven RAN monitor and Telenor-Ericsson’s capacity and power optimizer, highlight potential gains in reliability, latency, and cost. Juniper Research says true commercial offerings are still a year or more away. But did you know that BSS Magic from Totogi is a telco-specific AI agent-powered solution ready to connect to your entire BSS estate? Want to build some BSS agents? Give us a call!
Operators will invest more than $86 billion in AI in the next four years, according to Juniper Research. Why the big spend? Agentic AI, again! Beyond operational optimization and cost savings, telcos will also be using agentic AI for new revenue streams such as fixed-wireless access, private networks, and AI-enabled network slicing. With all this AI transformation we will start to see employee counts dropping to pay for all the investment. It’s about to get interesting…
CAMARA just issued its second major bundle, the Spring25 Meta-Release, adding 13 new and 23 updated APIs for developers to tap core network functions such as Quality of Service, SIM-swap checks, location, and more. The project now includes 1,250 contributors from 427 organizations, and aligns with GSMA Open Gateway and TM Forum standards. Are the network APIs taking off? Are telcos generating the revenue they thought with this new approach? Shoot me a DM and let me know!
Phishers, beware! SK Telecom is finding fraudsters with an AI-enabled voice phishing prevention service in South Korea. The anomaly-detection service, built on the telco’s ScamVanguard technology, safeguards both SKT’s Adot personal assistant and Industrial Bank of Korea’s SurPass platform, flagging suspicious calls in real time and blocking fraudulent transfers. Sounds like a great application that’s providing real business value. Woohoo!
The EU will spend €20 billion to build three to five “AI gigafactories” in an effort to catch up to the US and China, reduce reliance on foreign technology, and catch the next wave of AI innovation. The strategy also calls for using as much renewable energy as possible, developing homegrown AI chips, and working to simplify the recently passed AI Act. My money’s still on the private enterprises kicking the public sector’s asses on getting stuff done. Remember that big public cloud project called GaiaX that was supposed to compete with the hyperscalers? How’s that going? 🤦🏻♀️
But what if AI won’t use as much energy as we think? An AI startup just unveiled tools to run AI without the costly chips. India’s Ziroh Labs, in partnership with IIT Madras, has unveiled Kompact AI—a software stack that runs large models such as Llama 2 and Qwen 2.5 on ordinary computers (like an Intel Xeon laptop), eliminating the need for scarce, costly graphical processing units. Validated by Intel and AMD, the inference‑focused platform lowers the hardware barrier for Indian developers, dovetailing with the country’s ₹10,300‑crore IndiaAI Mission and broader push for tech self‑reliance. Maybe there’s an answer to the green paradox of AI after all…
This is what I mean about being AI-first: No new hires unless you can prove AI can’t do the job, per Shopify CEO Tobi Lütke. AI adoption has officially become the expectation across the company, and effective AI use will factor into performance and peer reviews. It’s coming from the top, and if Tobi is an effective leader, he should be able to drive it down into the rank and file. While this was the “CEO memo heard round the world,” it shouldn’t be news to Totogians. We’ve been thinking and working like this for years!
