← Back to PayloadsAI2026-04-13
S3 Files 📜, Bun for Pulumi 🥯, Caching at Netflix 📺
Amazon S3 Files is a new capability that allows any S3 bucket to be
mounted and accessed as a fully-featured file system
directly
Quick Access
Install command
$ mrt install ai

**TL;DR** - AWS S3 file handling improvements, Bun Pulumi integration, and how Netflix approaches caching at scale.
The 10-Second Pitch
- S3 flexible retrieval options let you trade cost for access speed with fine-grained control
- Bun Pulumi provider now production-ready - infrastructure as code in TypeScript without Node.js overhead
- Netflix caching architecture handles 250K+ requests per second with sub-millisecond P99 latency
Setup in 3 Steps
1. Evaluate S3 Intelligent-Tiering for data with unpredictable access patterns
2. Try Bun for your next Pulumi deployment script - startup time difference is noticeable
3. Study Netflix cache invalidation approach - their multi-layer caching is well-documented and applicable
**Example Prompt:**
Design a caching strategy for a read-heavy API handling 100K requests per minute with 95% cache hit rate.
Verdict
| Pros | Cons |
|---|
| S3 flexible retrieval genuinely useful | Requires careful monitoring to avoid bill shock |
| Bun + Pulumi is a real DX improvement | Bun ecosystem still has gaps vs Node |
|---|
| Netflix cache architecture well-documented | Replicating it requires significant operational infra |
Netflix cache documentation worth reading even if you are not Netflix - principles apply at smaller scales.