Kinyarwanda
Oshiwambo
Tagalog
Punjabi
Government institutions hear from citizens through two very different channels. One is direct: the complaints, questions and disputes that arrive through official lines. The other is ambient: what people say publicly, on social platforms and in the press, often long before they file anything formal. Most institutions act on the first and miss the second – until a localised grievance has already become a national one.
Proto’s Perception module closes that gap. It monitors what the public is saying across social media, app stores and news sites, categorises it, and analyses sentiment, so emerging issues surface while they can still be addressed rather than only after they escalate.
What Perception does
Perception scrapes posts, comments and reviews from Facebook, X, Instagram, YouTube, TikTok, news websites, Google Play and the Apple App Store, on a schedule the institution controls. Keyword groups – an agency’s name, a programme, a payment scheme, a known scam term – determine what is captured, so the data stays focused on what matters and avoids unrelated noise.
Every captured entry is automatically categorised and scored for sentiment. A dashboard visualises total volumes, the most-mentioned entities, category breakdowns and sentiment trends over any period, segmented by source. Smart alerts notify teams by email when a keyword is detected, when mention volumes cross a threshold within a time window, or when the share of negative sentiment exceeds a defined limit.
Why it matters for government
The value is not monitoring for its own sake – it is connecting public signal to operational response. Perception feeds directly into Proto’s Analytics module, where public sentiment can be compared against livechat and ticketing data in a single view. An institution can see a spike in public concern about a particular scam typology or service failure, correlate it with what is arriving through its AI agent channels, and act before the issue compounds.
For a regulator or public agency already running an AI agent for citizen engagement or consumer protection, Perception adds the missing half of the picture: not only resolving what citizens bring to the institution, but understanding what they are saying everywhere else.