graze-social

graze-social / personalization

Public

Graze Personalization Engine

18
1
100% credibility
Found Mar 10, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

A Rust microservice that generates personalized Bluesky feeds by analyzing like graphs with random walk algorithms and caching results in Redis.

How It Works

1
📱 Discover Graze

You hear about Graze, a tool that creates smart, personalized feeds for Bluesky based on what people like.

2
🔗 Connect likes

Link your Bluesky likes so it learns from real interests and builds connections between similar people.

3
🎨 Design your feed

Pick a style for your feed, like discovery for fresh ideas or stable for reliable favorites.

4
🚀 Launch it

Turn on your feed generator and watch it start recommending posts instantly.

5
⭐ Add highlights

Pin special posts or boost favorites to make your feed even more exciting.

😊 Perfect feeds

Your Bluesky friends love seeing posts that match their tastes perfectly.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 18 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is personalization?

Graze Personalization Engine is a Rust microservice that ranks Bluesky feeds in real-time using users' like history and similar users' tastes via a graph-based random walk. It solves the cold-start problem for custom feeds on the AT Protocol by blending personalized picks with fallbacks like trending or popular posts. Devs get a ready API (POST /v1/personalize), Docker Compose for quick starts, and optional ClickHouse for analytics—no database required initially.

Why is it gaining traction?

Its inverted algorithm precomputes co-likers for sub-100ms responses under load, with presets like "discovery" for niche content or "fast" for speed. Pluggable sources (HTTP, ClickHouse) and admin endpoints for candidates/special posts make iteration easy, plus diagnostics like feed status and Redis inspection. Rust + Axum delivers low-latency without Lua bottlenecks, standing out from slower LLM personalization github alternatives.

Who should use this?

Bluesky feed generator devs at graze social github building algorithmic timelines, AT Protocol startups needing like-graph ranking, or indie hackers prototyping personalized social apps. Ideal for teams handling Jetstream likes and wanting K8s-ready deploys without Salesforce-scale infra.

Verdict

Promising engine for Bluesky personalization with strong docs and Docker/K8s support, but 18 stars and 1.0% credibility signal early maturity—test coverage looks solid yet unproven at scale. Grab it if you're in the AT ecosystem; otherwise, monitor for production stability.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.