Similarity Search API Batch Processing

[evgenydmitriev]

We have a Similarity Search API built with Cloudflare Workers AI and Cloudflare Vectorize database. It's a simple Cloudflare worker that looks up incoming messages in a vector database and returns a similarity score. It's used by other services for near-duplicate detection, topic classification, and synthetic data generation purposes, among other things.

To participate in the challenge, submit a pull request that adds batch message processing to the Similarity Search API worker. Make sure that your solution doesn't introduce any new bottlenecks or increase overall costs of the Similarity Search API under load. Consider possible Cloudflare workers limitations, as well as cost and performance trade-offs. Focus on leveraging Cloudflare services' native functionality, rather than implementing needed functionality with custom code. Keep in mind that your submission will be evaluated based on the combination of resource efficiency, security, and speed. When done, request a review from this issue assignee. Expanding the pull request description with your methodology can help us better understand your reasoning and evaluate your submission faster.

Anyone can participate in this challenge. You don't need anyone's approval to try! All submissions will be reviewed by the maintainers, and additional changes to your pull request may be asked of you to bring your submission to improve code quality, security, and/or efficiency. See more details on the Challenge Program and check out the success stories of the challenge winners.

There are different ways of achieving the goal of the challenge, so don't get discouraged if someone already opened a pull request. Multiple bounties are available! We'll keep the challenge open until we get enough high-quality submissions. Looking forward to your pull requests!

How to Claim Bounty

Email your bitcoin or stablecoin payment address to challenge.bounties@dn.institute with a link to your merged pull request. Expect all payouts to be completed by the end of the month.