🕸️ DisSysLab — Build distributed apps by connecting functions

DisSysLab (aka dsl) is a Python framework for building distributed programs. A program is represented by a directed graph in which each node is a Python callable. You do not write threads, locks, or send/receive calls. The dsl runtime runs the nodes concurrently and moves messages along edges.

dsl introduces first-year undergraduates to distributed programs. Each student builds distributed-system applications that interests that specific student. You build systems by connecting functions in widely used libraries, such as NumPy, and LLM APIs such as OpenAI. dsl is an early release; it will evolve, and feedback is welcome.

The core idea

Edges carry messages (often Python dicts) from node to node. We call a node of the graph an agent. You specify a graph by its list of edges.

DisSysLab uses two main callable shapes:

1. Sources: yield a stream of messages
2. Transforms / Sinks: take one message and return one message (transform) or return nothing/ignored (sink)

Example: a graph g with three agents—data_source, data_transformer, and data_sink, and data_sink and two edges: (1) data_source to data_transformer and (2) data_transformer to data_sink.

from dsl import network

g = network([
    (data_source, data_transformer),
    (data_transformer, data_sink),
])

You can write agents as plain Python functions, or use dsl connectors/wrappers:

• Sources: sensors, RSS feeds, files, synthetic generators
• Transforms: library calls, filters, ML, LLM calls
• Sinks: console displays, JSONL/CSV recorders, databases, actuators

You can create arbitrary graphs. Early modules use acyclic graphs; later modules introduce cycles.

TL;DR – try it

git clone https://github.com/kmchandy/DisSysLab.git
cd DisSysLab
python -m venv .venv && source .venv/bin/activate   # Windows: .\.venv\Scripts\Activate.ps1
python -m pip install -e .
python -m modules.ch01_networks.simple_network

👉 Start Here

Make sure that you have the keys you need for your services (eg OpenAI) in your .env file for modules 3 and 5. Start with Module 1 — Intro → modules/ch01_networks/README_1.md

Modules

1. Module 1 — Intro → modules/ch01_networks/README_1.md
2. Module 2 — Sources → modules/ch02_sources/README_1.md
3. Module 3 — OpenAI agents → modules/ch03_GPT/README_1.md
4. Module 4 — Numerics → modules/ch04_numeric/README.md
5. Module 5 — Use AI to build systems → modules/ch05_use_AI/README.md