Rate limiting and budget controls in Agentic AI - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Rate Limiting Purpose

Why is rate limiting important in managing AI agent requests?

A To reduce the accuracy of AI model predictions intentionally

B To allow unlimited requests without any restrictions

C To prevent system overload by controlling the number of requests in a time frame

D To increase the speed of all incoming requests simultaneously

💡 Hint

Think about what happens if too many requests come at once.

Rate limiting helps avoid system crashes by limiting how many requests can be processed in a given time.

❓ Model Choice

intermediate

Choosing a Budget Control Strategy

You want to limit the total compute cost of AI agent calls over a month. Which budget control method is best?

A Set a fixed maximum number of API calls allowed per month

B Disable rate limiting and monitor usage manually

C Allow unlimited calls but slow down the response time

D Increase the model size to reduce calls

💡 Hint

Think about how to keep costs predictable.

Setting a fixed maximum number of calls directly controls usage and cost.

❓ Predict Output

advanced

Output of Rate Limiting Code Snippet

What is the output of this Python code simulating a simple rate limiter?

Agentic AI

import time

class RateLimiter:
    def __init__(self, max_calls, period):
        self.max_calls = max_calls
        self.period = period
        self.calls = []

    def allow(self):
        current = time.time()
        self.calls = [call for call in self.calls if call > current - self.period]
        if len(self.calls) < self.max_calls:
            self.calls.append(current)
            return True
        return False

limiter = RateLimiter(3, 5)
results = []
for _ in range(5):
    results.append(limiter.allow())
    time.sleep(1)
print(results)

A [False, False, False, False, False]

B [True, True, True, True, True]

C [True, False, True, False, True]

D [True, True, True, False, False]

💡 Hint

Consider how many calls are allowed within the 5-second window.

The limiter allows 3 calls in 5 seconds. The first 3 calls return True, the next 2 return False.

❓ Metrics

advanced

Interpreting Budget Control Metrics

You have a budget control system that tracks 'calls_used' and 'calls_remaining'. If calls_used = 750 and calls_remaining = 250, what is the total budget and percentage used?

A Total budget is 1000 calls; 75% used

B Total budget is 1000 calls; 25% used

C Total budget is 500 calls; 150% used

D Total budget is 1000 calls; 50% used

💡 Hint

Add used and remaining calls to find total, then calculate percentage.

Total budget = calls_used + calls_remaining = 1000. Percentage used = (750/1000)*100 = 75%.

🔧 Debug

expert

Debugging Budget Control Logic

Given this budget control snippet, what error will occur when running it?

Agentic AI

class BudgetControl:
    def __init__(self, max_calls):
        self.max_calls = max_calls
        self.calls_made = 0

    def make_call(self):
        if self.calls_made < self.max_calls:
            self.calls_made += 1
            return True
        else:
            return False

budget = BudgetControl(3)
results = [budget.make_call() for _ in range(5)]
print(results)
print(budget.calls_remaining)

A IndexError: list index out of range

B AttributeError: 'BudgetControl' object has no attribute 'calls_remaining'

C TypeError: unsupported operand type(s) for +: 'int' and 'str'

D No error; output is [True, True, True, False, False]

💡 Hint

Check if all attributes used are defined in the class.

The class does not define 'calls_remaining', so accessing it causes AttributeError.