FalconX Software Engineer Interview Experience — OA on the Longest Risk-Limited Trade-Log Window

Build an audit module for a high-frequency trading platform that processes an asynchronous, out-of-order data stream containing trade execution logs from multiple exchanges. The task is to reconstruct the chronological execution timeline for a specified exchange (target_venue), and find the maximum number of consecutive trade logs in a contiguous execution window that satisfies specific operational risk limits.

You are given an array of log objects logs, a target exchange identifier target_venue, a maximum allowed price deviation V, and a maximum cumulative slippage budget S. Process and solve it according to the following steps:

1. Filter: select the log records whose venue equals target_venue.
2. Reconstruct: sort the filtered records by timestamp in ascending order.
3. Calculate Metric: for each log, compute its slippage cost: slippage_cost = abs(fill_price - benchmark_price) * volume.
4. Evaluate Window: in the sorted logs, find the length of the longest contiguous sub-sequence that satisfies both of the following conditions at the same time:
   • The difference between the max price and the min price in the window is no greater than V: max(fill_price) - min(fill_price) <= V
   • The sum of the slippage costs of all logs in the window is no greater than S: sum(slippage_cost) <= S
5. Return: return the length of the longest window that satisfies the conditions above (i.e., the number of logs). If no valid window exists, or no logs match after filtering, return 0.