Risk & communications · Bank Negara Malaysia
Reputational risk monitoring platform
A multi-source social media analytics platform integrating LinkedIn, Facebook, Instagram, and Twitter APIs with NLP sentiment analysis and automated controversy flagging for reputational risk.
4 platforms unified into one real-time sentiment and controversy-tracking pipeline
The problem
A central bank’s reputation moves in public conversation long before it shows up in a formal complaint. Communications and risk teams had no systematic way to see that building, so they found out about a brewing controversy the same way everyone else did: after it was already a headline.
What I built
- A multi-source ingestion pipeline pulling from LinkedIn, Facebook, Instagram, and Twitter REST APIs into a MySQL backend
- NLP sentiment analysis scoring public conversation as it comes in, instead of relying on manual monitoring
- Automated controversy flagging that surfaces spikes in negative sentiment early, so the response is proactive rather than reactive
Why it matters
Four disconnected feeds became one real-time view. Risk and communications teams could see reputational exposure building and act on it, backed by data instead of anecdote.