Real-time Bursa alerts, scored and triaged by AI
The problem
Bursa Malaysia publishes thousands of corporate announcements every month: quarterly results, insider trades, material contracts, dividend declarations, share buybacks. For retail investors, this is all public information. The problem is keeping up with it.
Most investors find out about material filings hours after the fact, often only after the stock price has already moved. The existing options (checking the Bursa website company by company, scrolling through broker app notifications, following noisy Telegram channels) all require manual effort and still miss things. By the time you've read through a dense 40-page quarterly report, the market has already reacted.
We wanted to prove that this entire workflow could be automated. That an AI system could read every filing as it drops, decide whether it matters, and tell you why, all before most investors have even noticed it was published.
What we built
Sang Tikam is a live product, running right now, that monitors all Bursa Malaysia announcements in real time and delivers scored, summarised alerts to investors via Telegram. You can try it free for 7 days.
Real-time scraping and processing
The system watches Bursa Malaysia's announcement feed continuously. When a new filing appears, it's picked up within minutes, not hours. Every announcement is ingested, categorised, and queued for analysis automatically.
No manual checking. No refresh buttons. The system never sleeps.
AI-powered scoring and summarisation
Each announcement is analysed by AI and assigned an impact score from -5 (bearish) to +5 (bullish). The system identifies the type of filing (earnings, insider trade, contract win, restructuring) and generates a plain-English summary of what happened and why it matters.
A 40-page quarterly report becomes a two-paragraph summary with a clear signal: is this good news or bad news, and how significant is it?
Filtered Telegram alerts
Users configure their own alert thresholds and watchlists. If you only care about high-impact announcements (score ≥ 3 or ≤ -3), you won't be bothered by routine filings. If you have a watchlist of 20 stocks, you only hear about those.
Alerts arrive in Telegram within minutes of filing, complete with the score, summary, category, and a link to the full announcement.
How it works under the hood
The scraper runs as a persistent process that polls Bursa Malaysia's public filing feed at short intervals. New announcements are deduplicated and stored, then passed to an AI pipeline that extracts structured data (company, filing type, key figures) and generates both the impact score and natural-language summary.
Telegram delivery is handled through the Bot API, with per-user preferences stored in a lightweight database. The entire system runs on cloud infrastructure with no manual intervention required. It's been processing announcements 24/7 since launch.
The results
- 17,000+ announcements analysed across 900+ listed companies
- Alerts within minutes from filing to Telegram
- Noise reduced by ~80% because users only see announcements that meet their threshold
- Zero manual intervention. The system runs autonomously around the clock
- Live and available now. Sign up for a free 7-day trial and see it in action