aleeve.dev / ott
Scalable and distributed custom feed generator, ott - on that topic
fork atom
Rust 69.4%
Smarty 18.3%
Dockerfile 7.3%
Shell 1.5%
Other 3.5%
37 1 0

Clone this repository

https://tangled.org/aleeve.dev/ott
git@tangled.org:aleeve.dev/ott

For self-hosted knots, clone URLs may differ based on your setup.

Download tar.gz
README.md

OTT on that topic#

OTT is an over the top bluesky custom feed (in the making)

The flow is as follows:

  1. Two fluvio connector services consume wss streams, one for raw-posts and one for raw-likes.
  2. ott-filter consumes the keyed posts and likes streams, and keeps count on likes and other filters. It sends the passing posts to the fluvio topic posts.
  3. ott-embed consumes the posts topic, embeds them with tei running on host and stores the vectors in a pg cluster
  4. ott-xrpc listens to getFeedSkeleton requests, gets the users last liked post and gets similar posts from the pg db.

Still work in progress, especially the ott-xrpc service isn't fleshed out yet. Also I intend to add a VIP stream so that all posts liked by a feed user are guaranteed to pass the filter.

Then... The original intention was to use FASTopic to get topic vectors for every day in the semantic space and sample within the most relevant topics. This is still the intention.

Setup#

Note that the ott-helm currently assumes that you intend to deploy through a cloudflare zerotrust tunnel, and defaults to starting up a cloudflared daemon forwarding to the ott-xrpc service. If your DNS isn't configured like that and you have no tunnel simply disable cloudflare in your helm values.

Also note that this should likely not be done at all just yet, this software is in early development.

Install dependencies#

# Install k8s tooling
brew install kind helm skaffold

# Install fvm and fluvio cli
curl -fsS https://hub.infinyon.cloud/install/install.sh | bash

# Install rust
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh

# Install tei
cargo install --git https://github.com/huggingface/text-embeddings-inference

Create a cluster#

kind create cluster --config kind-cluster.yaml

Run#

skaffold run