Pathery :fire: Serverless Search :fire:

Pathery is a serverless search service built on AWS using Rust, CDK and Tantivy. It uses AWS managed serverless offerings – DynamoDB, EFS, Lambda, SQS, and API Gateway – to the maximum extent possible.

:bell: WARNING: This is currently a work in progress and not ready for production usage.

Features

• 🔥 Fast full-text search. Built on Rust to limit AWS Lambda cold start overhead.
• 🥰 Simple REST API. A simple REST API to make search as easy as possible.
• 👍 Easy to install. Ships as a CDK Component, making it easy to get started.
• 💵 Usage based infra. No long running servers, only pay for what you use.
• 🔼 Built for AWS. Leans on AWS managed services to limit maintenance burden and maximize scalability.
  • Document store: DynamoDB
  • Index store: Elastic File System (EFS)
  • Index writer & handler: Lambda
  • Index queue: SQS
  • API: API Gateway

Getting Started

Check out the getting started guide to deploy Pathery into your AWS account using CDK.

Architecture

Follow along with the Dev Log:

• Pathery Dev Log #1: Performant Serverless Queries Without a Cluster
• Pathery Dev Log #2: Indexing and the Document Store