assignment-8/video-script.txt at master · finm-python-for-finance/assignment-8 · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Title: Interprocess Trading Stack Walkthrough
Duration Target: ~4 minutes
Suggested Recording Setup: 4 PowerShell windows tiled (Gateway, OrderBook, Strategy, OrderManager) plus a Notepad window with performance_report.md for closing remarks. Capture with OBS at 1080p/30fps. Narrate live using the script below.

---

0:00 – 0:20  Introduce the project
- Say: “Hi, this is the assignment 8 trading stack. We have four independent Python processes communicating via TCP sockets and shared memory.”
- Briefly point to each terminal window: Gateway (top-left), OrderBook (top-right), Strategy (bottom-left), OrderManager (bottom-right).

0:20 – 1:10  Start supporting services
- In the OrderManager window, run `python order_manager.py`. Mention it listens for TCP orders and logs executions.
- In the OrderBook window, run `python orderbook.py`. Explain that it connects to the price feed and writes the latest prices into shared memory.
- Talk through how shared memory is recreated on start and that the strategy will attach later.

1:10 – 2:10  Launch the Gateway
- In the Gateway window, run `python gateway.py`.
- Highlight the log message confirming price/news clients will connect.
- Mention that prices are generated with a random-walk per symbol and news sentiment is random 0–100 values, all serialized with the shared delimiter.
- Point at the OrderBook terminal showing a successful connection and shared-memory updates.

2:10 – 3:10  Launch the Strategy
- In the Strategy window, run `python strategy.py`.
- Describe how it reads the shared-memory buffer, listens to the news stream, calculates short/long moving averages (3 vs 6 ticks), and only trades when price and sentiment agree.
- Wait for a logged trade in the OrderManager window. Call out the order fields, especially the embedded latency figure.

3:10 – 3:40  Demonstrate resiliency (optional clip)
- Stop the Gateway process with `Ctrl+C`. Narrate how OrderBook reports the loss and retries the connection.
- Restart the Gateway (`python gateway.py`) and point out how all other processes recover automatically and orders resume.

3:40 – 4:00  Wrap-up with documentation
- Switch briefly to Notepad showing `performance_report.md`. Summarize throughput (≈2 ticks/sec), decision latency (~40 ms), and reconnection behaviour.
- Close by mentioning `pytest` tests (`python -m pytest`) and that configuration lives in `config.py` for easy tuning.

---
Recording Tips:
- Keep microphone close to avoid room echo.
- If a section produces no orders immediately, remind viewers that sentiment must align with price – let the logs run for a few seconds.
- Trim any dead air at the start/end before exporting the final video.