← Latest papers
🤖 machine learning

Trackly: A Unified SaaS Platform for User Behavior Analytics and Real Time Rule Based Anomaly Detection

Trackly is a unified, scalable SaaS platform built on a multi-tenant microservices architecture that integrates comprehensive user behavior analytics with real-time, rule-based anomaly detection to simultaneously optimize digital experiences and mitigate security threats like fraud and bot attacks.

Original authors: Md Zahurul Haque, Md. Hafizur Rahman, Yeahyea Sarker

Published 2026-02-02
📖 4 min read☕ Coffee break read

Original authors: Md Zahurul Haque, Md. Hafizur Rahman, Yeahyea Sarker

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you run a busy digital store. You have two main problems to solve:

  1. The Shopkeeper's Dilemma: You want to know what your customers like, how long they stay, and where they get stuck so you can make the store better.
  2. The Security Guard's Dilemma: You need to catch thieves, bots, and people pretending to be someone else before they steal your data or money.

Usually, businesses hire two different teams for this. One team uses fancy tools to watch customer habits (Analytics), and another team uses different tools to watch for bad guys (Security). The problem? These two teams don't talk to each other. The security guard might not know that a "strange" login is actually just a loyal customer traveling for vacation, or the shopkeeper might miss a thief because they are too focused on sales charts.

Enter "Trackly."

Think of Trackly as a super-smart, all-in-one security guard who is also a master shopkeeper. It's a single platform that does both jobs at the same time, speaking the same language.

How Trackly Works (The "Magic" Behind the Scenes)

1. The Digital ID Card (Fingerprinting)
When someone visits your site, Trackly doesn't just look at their name. It creates a unique "digital fingerprint" of their device and browser. It's like checking a person's face, voice, and the specific way they walk. If someone tries to log in from a brand new device that doesn't match their usual "walk," Trackly raises an eyebrow.

2. The "Impossible Travel" Detector
This is one of Trackly's coolest tricks. It uses a math formula (called the Haversine formula) to measure how fast a person is moving between logins.

  • The Analogy: Imagine a customer logs in from New York at 9:00 AM and then logs in from London at 9:15 AM.
  • The Logic: Even a super-fast jet can't make that trip in 15 minutes. Trackly knows this is physically impossible. It flags this immediately as a potential thief using a stolen password, because no human could travel that fast.

3. The "Suspicious Activity" Scorecard
Trackly doesn't just say "Yes" or "No." It acts like a judge giving points for suspicious behavior.

  • Logging in from a new country? +2 points.
  • Using a "mask" (like a VPN) to hide their location? +1 point.
  • Clicking buttons 50 times in one second (like a robot)? +2 points.
  • Logging in at 3 AM when they usually sleep at 10 PM? +1 point.

If the score gets too high, Trackly sounds the alarm. If the score is low, it lets the user pass. This makes the system very clear about why it flagged someone, unlike some "black box" AI systems that just guess.

4. The Live Control Room (Dashboard)
Trackly gives the business owner a big screen showing everything happening right now.

  • The Map: You can see dots popping up all over the world showing where users are.
  • The Stats: It tells you how many people are active today, how long they stay, and how many are bouncing off the site.
  • The Red Flags: It highlights the suspicious sessions in red so you can deal with them instantly.

What Did They Prove?

The creators of Trackly tested it with a massive amount of fake data (simulating both good customers and bad hackers). Here is what they found:

  • Accuracy: It was right 98.1% of the time.
  • False Alarms: It rarely cried wolf. Only 2.25% of the time did it think a good customer was bad.
  • Speed: It caught the "impossible travel" and "robot" attacks almost perfectly.

The Bottom Line

Trackly is a tool designed for small to medium businesses that can't afford to buy expensive, separate security and analytics software. It unifies them into one easy-to-use system. It helps you understand your customers while simultaneously acting as a vigilant guard against fraud, bots, and account theft.

What it is NOT (based strictly on the paper):

  • It is not a magic crystal ball that predicts the future.
  • It does not currently use complex "black box" AI that you can't understand; it uses clear, explainable rules.
  • It is not yet a full mobile app tracker (it focuses on web/desktop for now).
  • It hasn't been tested in a real-world hospital or bank yet; the results come from simulated tests.

In short, Trackly is a practical, transparent, and efficient way to keep your digital house safe while learning how to make it more welcoming for your guests.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →