TECNO's EllaClaw, the new agentic AI built into the CAMON 50 Ultra, promises to reduce the everyday mental load of smartphone use by acting on intent rather than waiting for commands. After two days of testing, the AI proved genuinely useful for tasks like SMS triage, travel preparation, and data management, though it still has gaps in cross-app intelligence and autonomous action.
First Impressions: A Different Kind of Assistant
EllaClaw introduces itself proactively, asking what to call you and outlining its capabilities. Unlike traditional voice assistants that respond to commands, EllaClaw is designed as an agent that acts on goals. For example, instead of saying "set my alarm for 6am," you can say "I have an early flight tomorrow, help me not miss it," and the AI coordinates multiple steps. This shift from reactive to proactive interaction sets a new tone for smartphone AI.
The SMS Problem: Taming the Inbox
For Nigerian users, the SMS inbox is cluttered with bank alerts, OTPs, promotions, and broadcast messages. When asked to "summarise my important messages from this week, focus on anything financial," EllaClaw grouped messages by type—debits, credits, OTPs—and flagged a low-balance alert and a suspicious debit. It also noted expired OTPs that could be ignored. This saved minutes of scrolling and reduced distraction. However, the AI occasionally mistook promotional messages for genuine alerts, such as a loan offer dressed as a transaction notification. This gap in accuracy means users cannot fully rely on it for financial triage without verification.
Work Trip Test: One Sentence, Multiple Steps
When given the request "I'm travelling to Abuja for work on Thursday. Two meetings. Help me prepare," EllaClaw located calendar entries, organised an itinerary, pulled weather data, and suggested an alarm time. However, when a calendar entry lacked details like location or notes, the AI paused for clarification rather than making assumptions—a reasonable but disruptive behaviour. The AI also failed to cross-reference an SMS from an airline about a schedule change, missing a connection that would have made the trip planning more seamless. This highlights that the agentic AI's power grows with the richness of data in calendar, contacts, and notes.
Data Usage Diagnosis: Practical and Unglamorous
One unexpected standout feature was data management. When asked "My data is running out, and I have five days until renewal. What's eating it?" EllaClaw pulled a breakdown of data usage by app, revealing that a video app and a social media app were consuming data in the background. It offered step-by-step guidance to turn off background data, but did not execute the change autonomously—a consistent autonomy boundary. For markets where mobile data is expensive, this feature solves a real, everyday problem without fanfare.
The Honest Assessment: What EllaClaw Is and Isn't
EllaClaw is genuinely useful for reducing friction in daily tasks like SMS triage, data checks, and loose travel prep. However, it is not yet fully autonomous as marketing suggests. The 'one sentence, multiple steps' promise works best when the phone's environment is well-organised—detailed calendar entries, clean contacts, and labelled notes. Cross-app intelligence, where calendar, messages, and context connect without prompting, is still evolving. TECNO positions EllaClaw as "Practical AI" rather than performative AI, focusing on reducing cognitive load rather than showcasing clever tricks. According to the article, "the ambition is much more useful than that. It's to build a phone that quietly handles the low-level cognitive load of everyday life."
After two days, the agent phone era feels close but not fully realised. EllaClaw represents a significant step toward a smartphone that acts as an intelligent partner rather than a collection of apps. For now, it earns its place as a helpful tool, not a magic solution.
