What it is
Stelvio is an open-source Python framework that abstracts AWS infrastructure complexity - think Lambda, API Gateway, DynamoDB - into pure Python code and a CLI. Developer Michal Martinka's pitch: deploy serverless apps with stlv deploy instead of wrestling with Terraform or CDK YAML.
The project uses Pulumi under the hood but hides it. Auto-generates IAM roles, handles Python dependency packaging, and includes a dev mode for local Lambda testing against real AWS infrastructure. Apache 2.0 licensed.
The timing question
This launch comes as Python Lambda developers face familiar pain points. Container image optimization remains manual work - multi-stage Dockerfiles, layer caching, fighting the 10GB size limit. Cold starts still matter despite SnapStart (which doesn't support Python anyway). CDK Python support exists but developers consistently call it verbose compared to TypeScript.
Stelvio's 201 GitHub stars (3 forks) put it nowhere near SST's 15k+ traction. That's fine for a solo project - but it matters for production bets.
What to watch
The author was clear 54 days ago: v0.1 was "simple APIs only." Current version (v0.5) is nearing beta with stable core but evolving APIs. The Hacker News thread from this week shows expected skepticism around maturity versus established tools.
Three things matter here:
Python-first focus: Unlike SST (JS/TS-centric) or CDK (multi-language but verbose), Stelvio optimizes for Python workflows. Whether that's differentiation or limitation depends on your stack.
Dev mode specifics: Local testing against AWS infrastructure is interesting if it actually works. The "live code changes without redeployment" claim needs verification.
Production timeline: No clear signal on when the author considers this production-grade. Early adopters should plan accordingly.
The real question
Does the APAC Python serverless community need another abstraction layer? CDK works. Pulumi works. Terraform works. They're just not optimized for Python developer experience.
If Stelvio meaningfully reduces time-to-first-deploy without introducing new complexity, it fills a gap. If it's Pulumi with extra steps, it doesn't. We'll see which it turns out to be.
No funding disclosed. Solo maintainer. Explicitly pre-production. Interesting approach, early days.