# The languages

Four, and each one is here for a job the others do worse. For every one: **why I use it, how long I have written in it, and what I actually build with it**.

> mitsuki.cc/stack · this is the Markdown twin of the page at that URL.
> previous: mitsuki.cc/projects.md · step 3 of 8 · next: mitsuki.cc/stack/storage.md

## Rust

**Status:** In active use
  
**Site:** https://www.rust-lang.org

Main language. I reach for it where predictability and low latency matter: high-load services, data processing, platform cores.

- **Experience:** 3 years
- **Role:** primary

**Libraries and frameworks**

- Web / HTTP: Actix Web, Axum, Hyper
- Async: Tokio, Futures, async-trait
- Serialization / RPC: Serde, serde_json, rkyv, Bincode, Cap’n Proto
- Database: Diesel, ScyllaDB Rust Driver, Redis
- Networking: Tonic, Prost, Tungstenite
- Bots: teloxide
- Errors / observability: anyhow, thiserror, tracing, tracing-subscriber

Full page: mitsuki.cc/stack/rust.md

## Kotlin

**Status:** In active use
  
**Site:** https://kotlinlang.org

Came in for one specific project and stayed. I use it for JVM services and tasks where development speed matters without giving up strict types.

- **Experience:** 1 year
- **Role:** second primary

**Libraries and frameworks**

- Backend: Ktor
- Async: Kotlin Coroutines, Kotlin Flow
- Serialization: kotlinx.serialization
- Database: Exposed

Full page: mitsuki.cc/stack/kotlin.md

## Elixir

**Status:** In use
  
**Site:** https://elixir-lang.org

I reach for it where massive concurrency and tolerance to partial failure matter: the BEAM holds tens of thousands of lightweight processes, and one crashing doesn't take the system down.

- **Where:** microservices, concurrency

**Libraries and frameworks**

- Web: Phoenix, Plug
- Concurrency / OTP: OTP, GenServer, Supervisor
- Database: Ecto, Postgrex
- HTTP: Req, Finch

Full page: mitsuki.cc/stack/elixir.md

## Python

**Status:** Task-specific
  
**Site:** https://www.python.org

When getting the result fast matters more than running fast: prototypes, one-off data processing, scripts.

- **Experience:** ~2 years
- **Role:** supporting

**Libraries and frameworks**

- Web: Flask
- Async / HTTP: asyncio, aiohttp, httpx
- Data / scientific: NumPy, pandas, Polars
- AI / ML: PyTorch, Transformers
- Validation: Pydantic
- Testing: pytest
- Tooling: Ruff, uv

Full page: mitsuki.cc/stack/python.md