Free, no AI, runs in your browser
Cost lab: build a cloud bill
Interviewers ask “what would this cost a month, and how would you cut it?” Pick services for a real-looking workload, read the bill line by line, and get it under budget without breaking the requirements.
Prices are list prices for AWS Mumbai (ap-south-1), checked 19 Oct 2026, and may change. Use the numbers to build intuition and show your arithmetic; they are not a quote.
Pick a workload
0 of 8 solved- CoreBudget$2,575 a month5 cost levers
Food delivery in 3 Indian cities
200k orders a day, dinner peaks at 8 times the average, 2 TB of menu photos.
Starting bill: Compute 48%, Network 20%
- StretchBudget$3,550 a month4 cost levers
UPI-style payments API
5M payments a day, steady all day, p99 under 150 ms, logs kept 90 days.
Starting bill: Database 54%, Monitoring 22%
- SeniorBudget$3,500 a month4 cost levers
IPL live score and match metadata
Millions refresh the same score. Peak is 14x the average. Logs are huge.
Starting bill: Compute 74%, Database 17%
- StretchBudget$7,750 a month5 cost levers
E-commerce sale month
A flash sale at 8 pm jumps traffic 12x in minutes. Autoscaling alone is too slow.
Starting bill: Compute 81%, Database 9%
- CoreBudget$1,650 a month5 cost levers
B2B SaaS analytics dashboard
Busy 9 to 6 on weekdays, quiet at night. A 99.5% promise, not 99.99%.
Starting bill: Database 47%, Compute 41%
- SeniorBudget$8,600 a month3 cost levers
Photo-sharing app
120 TB of photos, 90 TB a month out to phones. Bytes are the bill.
Starting bill: Network 55%, Storage 24%
- CoreBudget$2,250 a month4 cost levers
Notification service
80M notifications a day through a queue. Campaign blasts run 10x. The queue itself costs money.
Starting bill: Compute 39%, Messaging 30%
- StretchBudget$3,400 a month3 cost levers
AI chatbot feature (LLM token costs)
3M chats a month on a large model. The tokens are the bill, not the servers.
Starting bill: AI model 99%, Compute 1%