Briefly describe the conceptual approach you chose! What are the trade-offs? The approach I have taken for the mentioned problem is as follow: 1. Divide the problem into sub problem first like a. Identifying Data Models of Rule Engine(Rule Class, Signal Class) based on requirements b. Building Utility Methods to validate rules 2. Development 3. Testing 4. Bug Fixing if any 5. Release
What's the runtime performance? What is the complexity? Where are the bottlenecks? Time Complexity of the solution is O(n). and runtime performance is also pretty well. CPU UTILIZATION and MEMORY UTILIZATION can become a bottlenecks here if the incoming data stream and rules is huge in size.
If you had more time, what improvements would you make, and in what order of priority? I have put only one day worth of work on it i.e 8 hours and prioritised the things needs i need to do. If I would have been allowed with extra time, 1. I would have planned to come up with an web application where user can manage their rules for the stream data and validate the stream data. 2. This was a beutifull problem that can be solved using Artificial Intelligence technolgies also by training with all such rules. if time would have permit me then i would have tried to solve it using Machine Learning algorithms.