Tell me about a time you had a strong technical disagreement with a Tech Lead, Architect, or peer.

In Simple Words: Demonstrate technical maturity: focus on data over ego, explore trade-offs respectfully, test hypotheses with prototypes, and practice disagree and commit.

The STAR Response:

• Situation: When rebuilding our user notification service, my Tech Lead wanted to adopt a complex distributed event streaming cluster (Apache Kafka), while I advocated for Amazon SQS + SNS, given our team size of 4 engineers and peak load of only 1,500 events/minute.
• Task: We needed to align on an architecture without stalling sprint velocity or creating interpersonal friction.
• Action:
  1. Avoided theoretical debates and organized an objective Trade-Off Matrix comparing operational burden, maintenance cost, cluster management, and latency.
  2. Built a rapid prototype demonstrating that SQS FIFO met 100% of our ordering and throughput requirements with zero infrastructure maintenance overhead.
  3. Acknowledged the Tech Lead's long-term scaling vision and agreed to establish clear metric triggers: if throughput exceeded 50,000 events/sec, we would re-evaluate Kafka.
• Result: The team unanimously agreed on SQS/SNS, shipping the feature 4 weeks faster than estimated. Two years later, the service still runs reliably at $25/month with zero operational maintenance incidents.

Practical Script & Summary Template

### Objective Architectural Trade-Off Matrix
| Criteria | Amazon SQS + SNS | Apache Kafka Cluster |
| :--- | :--- | :--- |
| **Operational Overhead** | Fully Managed (Zero maintenance) | ZooKeeper/KRaft node maintenance |
| **Cost at 1.5k msg/min** | ~$25 / month | ~$350 / month (cluster minimums) |
| **Throughput Ceiling** | 300 - 3,000 msg/sec (FIFO) | 1,000,000+ msg/sec |
| **Team Familiarity** | 4 of 4 engineers experienced | 1 of 4 engineers experienced |