Capstone

01-python-foundations/my_system_health.py

@@ -0,0 +1,32 @@
+import psutil
+
+try:
+    threshold = input("Enter the CPU threshold (%): ")
+    threshold = float(threshold)
+except ValueError:
+    print("That's not a valid number. Please enter something like 75 or 80.5")
+    exit()
+
+cpu_usage = psutil.cpu_percent(interval=1)
+memory_usage = psutil.virtual_memory().percent
+disk_usage = psutil.disk_usage("/").percent
+
+print("You entered:", threshold)
+print("Current CPU usage:", cpu_usage)
+print("Current Memory usage:", memory_usage)
+print("Current Disk usage:", disk_usage)
+
+if cpu_usage > threshold:
+    print("CPU status: WARNING - usage is above threshold")
+else:
+    print("CPU status: Healthy")
+
+if memory_usage > threshold:
+    print("Memory status: WARNING - usage is above threshold")
+else:
+    print("Memory status: Healthy")
+
+if disk_usage > threshold:
+    print("Disk status: WARNING - usage is above threshold")
+else:
+    print("Disk status: Healthy")

02-apis-and-json/call_api.py

@@ -2,7 +2,7 @@
 
 import requests
 
-API_URL = "https://jsonplaceholder.typicode.com/todos/1"
+API_URL = "https://jsonplaceholder.typicode.com/todos/2"
 
 
 def fetch_todo(url):
@@ -17,8 +17,8 @@ def main():
     for key, value in todo.items():
         print(f"{key:10}: {value}")
 
-    if todo.get("userId") == 1:
-        print("\n>> This todo belongs to user 1")
+    if todo.get("userId") == 2:
+        print("\n>> This todo belongs to user 2")
 
 
 if __name__ == "__main__":

02-apis-and-json/github_user.py

@@ -0,0 +1,24 @@
+import json
+import requests
+
+username = input("Enter a GitHub username: ")
+url = f"https://api.github.com/users/{username}"
+
+try:
+    response = requests.get(url, timeout=10)
+    response.raise_for_status()
+except requests.exceptions.HTTPError:
+    print(f"Could not find a GitHub user called '{username}'.")
+    exit()
+
+data = response.json()
+
+print("Name        :", data.get("name"))
+print("Public repos:", data.get("public_repos"))
+print("Followers   :", data.get("followers"))
+print("Location    :", data.get("location"))
+
+with open("github_user.json", "w", encoding="utf-8") as f:
+    json.dump(data, f, indent=2)
+
+print("\nSaved full response to github_user.json")

02-apis-and-json/stock_market_api.py

@@ -24,7 +24,7 @@ def get_daily_series(symbol, api_key):
 def main():
     api_key = os.environ.get("ALPHAVANTAGE_API_KEY")
     if not api_key:
-        print("Set ALPHAVANTAGE_API_KEY first:  export ALPHAVANTAGE_API_KEY=...")
+        print("Set ALPHAVANTAGE_API_KEY first:  export ALPHAVANTAGE_API_KEY=MCHYRHM0ME21PCKH")
         sys.exit(1)
 
     symbol = input("Enter a stock symbol (e.g. IBM, AMZN, GOOGL): ").strip().upper()

03-file-handling-and-logs/my_log_analyzer.py

@@ -0,0 +1,40 @@
+import json
+
+LOG_FILE = "app.log"
+
+try:
+    with open(LOG_FILE, "r", encoding="utf-8") as f:
+        lines = f.readlines()
+except FileNotFoundError:
+    print(f"Log file not found: {LOG_FILE}")
+    exit()
+
+print("Total lines:", len(lines))
+
+info_count = 0
+warning_count = 0
+error_count = 0
+
+for line in lines:
+    words = line.split()
+    if "INFO" in words:
+        info_count += 1
+    if "WARNING" in words:
+        warning_count += 1
+    if "ERROR" in words:
+        error_count += 1
+
+print("INFO   :", info_count)
+print("WARNING:", warning_count)
+print("ERROR  :", error_count)
+
+summary = {
+    "INFO": info_count,
+    "WARNING": warning_count,
+    "ERROR": error_count,
+}
+
+with open("log_summary.json", "w", encoding="utf-8") as f:
+    json.dump(summary, f, indent=2)
+
+print("\nSaved summary to log_summary.json")

04-object-oriented-python/log_analyzer_oop.py

@@ -9,7 +9,12 @@
 class LogAnalyzer:
     def __init__(self, log_file):
         self.log_file = log_file
-        self.counts = {level: 0 for level in LEVELS}
+        self.counts = {
+            "INFO": 0,
+            "WARNING": 0,
+            "ERROR": 0,
+            "UNKNOWN": 0,
+        }
 
     def read_logs(self):
         try:
@@ -32,7 +37,9 @@ def analyze(self, lines):
                 self.counts["UNKNOWN"] += 1
         return self.counts
 
-    def write_summary(self, path="log_counts.json"):
+    def write_summary(self, path=None):
+        if path is None:
+            path = Path(self.log_file).with_suffix(".json")
         with open(path, "w", encoding="utf-8") as f:
             json.dump(self.counts, f, indent=2)
 
@@ -47,6 +54,8 @@ def main():
         return
 
     result = analyzer.analyze(lines)
+    analyzer.write_summary("custom_log_counts.json")
+
     print("Log Analysis Summary:")
     for level, count in result.items():
         print(f"  {level:7}: {count}")

05-cli-tools-argparse/my_log_analyzer_cli.py

@@ -0,0 +1,52 @@
+import argparse
+import json
+import sys
+from pathlib import Path
+
+LEVELS = ("INFO", "WARNING", "ERROR")
+
+parser = argparse.ArgumentParser(description="Analyze a log file for INFO/WARNING/ERROR counts.")
+parser.add_argument("--file", required=True, help="path to the log file")
+parser.add_argument("--out", help="write the summary to this JSON file")
+parser.add_argument("--level", choices=LEVELS, help="show the count for only this level")
+
+args = parser.parse_args()
+
+log_path = Path(args.file)
+if not log_path.is_file():
+    print(f"Error: log file not found: {args.file}", file=sys.stderr)
+    sys.exit(2)
+
+with open(log_path, "r", encoding="utf-8") as f:
+    lines = f.readlines()
+
+info_count = 0
+warning_count = 0
+error_count = 0
+
+for line in lines:
+    words = line.split()
+    if "INFO" in words:
+        info_count += 1
+    if "WARNING" in words:
+        warning_count += 1
+    if "ERROR" in words:
+        error_count += 1
+
+counts = {
+    "INFO": info_count,
+    "WARNING": warning_count,
+    "ERROR": error_count,
+}
+
+if args.level:
+    print(f"{args.level}: {counts[args.level]}")
+else:
+    print("INFO   :", info_count)
+    print("WARNING:", warning_count)
+    print("ERROR  :", error_count)
+
+if args.out:
+    with open(args.out, "w", encoding="utf-8") as f:
+        json.dump(counts, f, indent=2)
+    print(f"Wrote summary to {args.out}")

05-cli-tools-argparse/my_summary.json

@@ -0,0 +1,5 @@
+{
+  "INFO": 10,
+  "WARNING": 2,
+  "ERROR": 3
+}

05-cli-tools-argparse/summary.json

@@ -0,0 +1,5 @@
+{
+  "INFO": 10,
+  "WARNING": 2,
+  "ERROR": 3
+}

06-aws-automation-boto3/cdk_demo/cdk.json

@@ -0,0 +1,3 @@
+{
+  "app": "python app.py"
+}

README.md

@@ -1,5 +1,9 @@
 # Python for DevOps + Agentic AI
 
+<p align="center">
+  <img src="https://github-readme-activity-graph.vercel.app/graph?username=elviDev&theme=github-light&cache_seconds=1800&v=3" alt="Elvis's Contribution Graph" width="100%" />
+</p>
+
 Learn to use Python for real DevOps work — automation, cloud operations,
 log analysis, internal tooling, and local AI agents.
 

capstone/README.md

@@ -6,6 +6,61 @@ interview-ready.
 
 ---
 
+## My implementation
+
+A FastAPI service — "DevOps Intelligence API" — covering the full "Impressive"
+tier: log analysis, system metrics, AWS resource visibility, and an
+AI-assisted log investigation endpoint backed by a local Ollama model.
+
+### Endpoints
+
+| Method | Path           | What it does                                      |
+| ------ | -------------- | -------------------------------------------------- |
+| GET    | `/health`      | Liveness check                                      |
+| GET    | `/logs`        | INFO/WARNING/ERROR counts for a log file            |
+| GET    | `/metrics`     | Current CPU, memory, and disk usage                 |
+| GET    | `/aws/s3`      | S3 buckets grouped by age                           |
+| GET    | `/aws/ec2`     | EC2 instances and their state                       |
+| GET    | `/aws/report`  | Combined S3 + EC2 inventory                         |
+| POST   | `/ai/analyze`  | Plain-English log summary from a local LLM agent    |
+
+### Setup
+
+```bash
+cd capstone
+python -m venv .venv
+.venv\Scripts\activate       # Windows
+# source .venv/bin/activate  # macOS/Linux
+
+pip install -r requirements.txt
+```
+
+### Run
+
+```bash
+uvicorn app.main:app --reload
+```
+
+The API is then available at `http://127.0.0.1:8000` (interactive docs at
+`/docs`). `/logs` and `/ai/analyze` default to `sample_logs/app.log` if no
+`file` is given.
+
+The `/aws/*` endpoints need AWS credentials available to `boto3` (e.g. via
+`aws configure` or environment variables). `/ai/analyze` needs a local
+[Ollama](https://ollama.com) instance running with the model set in
+`OLLAMA_MODEL` (defaults to `qwen3.6:latest`) pulled and available.
+
+### Tests
+
+```bash
+pytest
+```
+
+All 16 tests run without live AWS credentials or a running Ollama instance —
+AWS and AI calls are mocked.
+
+---
+
 ## The story these modules tell
 
 Every module in this course builds one capability. The capstone assembles them:

capstone/STAR.md

@@ -1,35 +1,46 @@
 # My Capstone — S.T.A.R Explanation
 
-Fill this in with your own words. A lot of people struggle in interviews not for
-lack of skills, but because they can't explain their work clearly. Bullet points
-are fine.
-
 ## Situation
-<!-- What was the problem or context? -->
-- 
+- Our team's application logs and infrastructure health were only checked
+  manually — someone had to SSH in, tail log files, and separately check AWS
+  and system metrics whenever something seemed off. That was slow, easy to
+  forget, and gave no consistent way to check status on demand.
 
 ## Task
-<!-- What were you responsible for? -->
-- 
+- I was responsible for automating this with Python: parse and summarize
+  application logs, expose system health metrics, surface AWS resource
+  inventory, and make all of it available through a single, reusable service
+  instead of one-off scripts.
 
 ## Action
-<!-- What did you actually build/do? Be specific about the tech. -->
-- 
+- Wrote a log parser that reads a log file and counts INFO/WARNING/ERROR
+  occurrences using whole-word matching, then wrapped it in an
+  `analyze_log_file` function returning structured counts.
+- Built a `psutil`-based metrics service reporting CPU, memory, and disk
+  usage, with a configurable CPU threshold that flags the system as healthy
+  or under high load.
+- Used `boto3` to build an AWS inventory service: S3 buckets grouped by age
+  (new vs. old, based on a configurable day threshold) and EC2 instances with
+  their current state.
+- Exposed all of it through a FastAPI service (`/health`, `/logs`, `/metrics`,
+  `/aws/s3`, `/aws/ec2`, `/aws/report`) with routers, services, and Pydantic
+  schemas kept in separate layers, and consistent HTTP error mapping
+  (404/400/500/502) for missing files, bad input, and AWS/API failures.
+- Added a local AI layer: a LangChain/LangGraph agent running against a local
+  Ollama model, given a tool that returns the exact deterministic log counts
+  so it can't invent statistics, exposed via `POST /ai/analyze`. Added
+  path-traversal protection so the endpoint can only read `.log` files inside
+  the approved log directory.
+- Wrote 16 unit tests covering the log parser, metrics service, AWS service
+  (mocking `boto3` clients so no real AWS calls are made in tests), and the
+  AI router's path-resolution and security checks.
 
 ## Result
-<!-- What changed? Quantify if you can (time saved, errors caught). -->
-- 
-
----
-
-### Example (for reference — replace with your own)
-
-- Situation: Application logs grew daily and manually scanning them for errors
-  was slow and error-prone.
-- Task: Automate log analysis with Python so issues surface quickly.
-- Action: Wrote a log parser that counts INFO/WARNING/ERROR, refactored it into
-  a class, added an `argparse` CLI, then exposed it via a FastAPI `/logs`
-  endpoint alongside `/metrics` and `/health`. Added a local LangGraph + Ollama
-  agent to summarize logs in plain English.
-- Result: Reduced manual log review effort and gave the team quick, on-demand
-  visibility into application health through a single API.
+- Replaced manual log/metrics/AWS checks with a single running API that
+  answers each question in one request instead of several manual steps.
+- The AI endpoint gives a plain-English read of what a log file shows,
+  grounded in the same deterministic counts the API reports elsewhere, which
+  makes it trustworthy enough to actually rely on rather than just a demo.
+- All 16 tests pass and the AWS/AI tests run without needing live AWS
+  credentials or a running Ollama instance, so the suite is safe to run in
+  CI.