Handle Rate Limits and Retries

Failures are normal

APIs return rate-limit (429) and transient errors. Production code must retry gracefully instead of crashing.

Exponential backoff

import time, random
def with_retries(fn, tries=5):
    for i in range(tries):
        try:
            return fn()
        except RateLimitError:
            time.sleep((2 ** i) + random.random())  # backoff + jitter
    raise RuntimeError("exhausted retries")

Also worth doing

Set timeouts, cap concurrency, and respect any Retry-After header. The jitter prevents many clients from retrying in lockstep and hammering the API at once.