How Power BI Helps You Land a Cloud Engineer Job Faster

# Monthly cost per service for a date range (AWS CLI v2).
# Cost Explorer must be enabled in the account first.
# Note: each Cost Explorer API request is billed (small charge).
aws ce get-cost-and-usage \
  --time-period Start=2026-08-01,End=2026-11-01 \
  --granularity MONTHLY \
  --metrics UnblendedCost \
  --group-by Type=DIMENSION,Key=SERVICE \
  --output json > cost_by_service.json

# Flatten the Cost Explorer JSON into a CSV for Power BI.
import csv
import json

with open("cost_by_service.json", encoding="utf-8") as f:
    data = json.load(f)

with open("cost_by_service.csv", "w", newline="", encoding="utf-8") as out:
    w = csv.writer(out)
    w.writerow(["month_start", "service", "cost_usd"])
    for period in data["ResultsByTime"]:
        start = period["TimePeriod"]["Start"]
        for g in period["Groups"]:
            amount = float(g["Metrics"]["UnblendedCost"]["Amount"])
            w.writerow([start, g["Keys"][0], round(amount, 4)])

print("Wrote cost_by_service.csv - load it with Get Data > Text/CSV")

-- Daily cost by service and team tag from a CUR table in Athena.
-- Column names follow the legacy CUR schema; CUR 2.0 uses
-- similar names but stores tags in a map column. Check yours.
SELECT
  date(line_item_usage_start_date)        AS usage_date,
  line_item_product_code                  AS service,
  COALESCE(NULLIF(resource_tags_user_team, ''), 'untagged') AS team,
  SUM(line_item_unblended_cost)           AS cost_usd
FROM cur_database.cur_table
WHERE line_item_usage_start_date >= date_add('day', -90, current_date)
GROUP BY 1, 2, 3
ORDER BY usage_date, cost_usd DESC;

-- Measures for a table named Cost with columns
-- usage_date, service, team, cost_usd, and a 'Date' table.

Total Cost = SUM ( Cost[cost_usd] )

Cost Last Month =
CALCULATE ( [Total Cost], DATEADD ( 'Date'[Date], -1, MONTH ) )

Cost Change % =
DIVIDE ( [Total Cost] - [Cost Last Month], [Cost Last Month] )

Untagged Cost % =
DIVIDE (
    CALCULATE ( [Total Cost], Cost[team] = "untagged" ),
    [Total Cost]
)

# Export 14 days of daily average CPU per running EC2 instance.
# pip install boto3   (uses your configured AWS credentials)
import csv
from datetime import datetime, timedelta, timezone
import boto3

ec2 = boto3.client("ec2", region_name="ap-south-1")
cw = boto3.client("cloudwatch", region_name="ap-south-1")
end = datetime.now(timezone.utc)
start = end - timedelta(days=14)

ids = [i["InstanceId"]
       for r in ec2.describe_instances(Filters=[{"Name": "instance-state-name", "Values": ["running"]}])["Reservations"]
       for i in r["Instances"]]

with open("ec2_cpu_daily.csv", "w", newline="") as f:
    w = csv.writer(f)
    w.writerow(["instance_id", "day", "avg_cpu"])
    for iid in ids:
        stats = cw.get_metric_statistics(
            Namespace="AWS/EC2", MetricName="CPUUtilization",
            Dimensions=[{"Name": "InstanceId", "Value": iid}],
            StartTime=start, EndTime=end, Period=86400, Statistics=["Average"])
        for p in sorted(stats["Datapoints"], key=lambda d: d["Timestamp"]):
            w.writerow([iid, p["Timestamp"].date(), round(p["Average"], 2)])

-- Table 'CPU' has instance_id, day, avg_cpu.
-- Flag instances whose average CPU stayed low over the period.

Avg CPU = AVERAGE ( CPU[avg_cpu] )

Idle Flag =
IF ( [Avg CPU] < 5, "Review: possibly idle", "In use" )

-- Put instance_id, [Avg CPU] and [Idle Flag] in a table visual.
-- The 5% threshold is only a starting point; agree real
-- thresholds with the team and check memory and network too.

PROJECT BRIEF: CLOUD COST AND USAGE DASHBOARD

Goal         : Show where cloud spend goes and what looks idle
Data sources : AWS Cost Explorer or CUR export (CSV/Athena)
               CloudWatch CPU export (boto3 script)
Model        : Cost fact, CPU fact, Date table, Team mapping
Pages        : Overview | By service | By owner tag | Idle review
Measures     : Total Cost, Cost Change %, Untagged Cost %, Avg CPU
Automation   : Script runs on schedule, files land in S3
Safety       : Budget alert on, read-only IAM role for exports
Repo         : README, screenshots, scripts, sample data

RESUME LINE (edit to match what you really did)

Built a Power BI dashboard on AWS cost and CloudWatch data
using Athena SQL and DAX to show cost by service and owner
tag and to list low-utilisation EC2 instances for review.

GITHUB README OUTLINE
1. Problem the dashboard answers
2. Architecture diagram: exports -> S3 -> Athena -> Power BI
3. IAM permissions used (read-only)
4. How to run the export scripts
5. Screenshots of each report page
6. What I would add next

INTERVIEW STORY (STAR format)

Situation: Cloud costs are hard to explain without a clear view.
Task     : Build a report showing spend by service and owner,
           plus resources that may be idle.
Action   : Exported cost data, wrote Athena SQL, built a date
           table and DAX measures, exported CPU metrics with
           boto3, and flagged low-usage instances for review.
Result   : A dashboard that answers why cost changed and who
           owns it. (State only what you actually observed.)
Learned  : Tagging matters; low CPU is a review signal, not proof.