ReactiveSearch is now completely free and open-source.
Search relevancy, pipelines, AI search, analytics, access control, caching — the full stack. No paywalls, no restrictions.
Get it running in under 5 minutes with Docker.
🧵↓
ReactiveSearch is now completely free and open-source.
Search relevancy, pipelines, AI search, analytics, access control, caching — the full stack. No paywalls, no restrictions.
Get it running in under 5 minutes with Docker.
🧵↓
Get started:
git clone https://t.co/H9aOzyHY5b
cd reactivesearch-api-docker
docker-compose -f docker-compose-with-elasticsearch.yaml up -d
That's it. Elasticsearch + ReactiveSearch API + Nginx + logging — one command.
@saranormous@ReactiveCo - Easiest way to deploy search in you app, 10x cheaper than algolia
@vercel - Defacto for React apps
@supabase - Postgres made easy
@AlchemyPlatform - Fast and Reliable RPC provider
@getsentry - error monitoring and tracking
@posthog - product analytics at 1/5 price
Introducing ReactiveSearch for React v4:
* AI Answer + Image Search support (Inferencing support choice b/w OpenAI's vision API and Google's Vertex AI),
* Embed charts in search UIs (20+ chart types supported with Apache E-Charts),
* FAQs, Featured Suggestions,
* Simplified SSR config (3x smaller),
* React v18 support,
* a secure API to express search intent, and
* expanding search engine support for Elasticsearch, OpenSearch, Solr, MongoDB and OpenAI.
👟 Quick start with developer guide: https://t.co/MYzrzg2NIU
ReactiveSearch aims to provide an incremental AI adoption strategy for your business search as an alternative to betting on new solutions.
If your site search uses Elasticsearch, OpenSearch or Solr, you *can choose* to improve it by bringing context awareness, AI answering capabilities, always returning a result and more with ReactiveSearch pipelines and Search UI.
Let's see this with a closer look at building an e-commerce search pipeline that does all of the above:
https://t.co/U1VW3xujjv
How to build a context aware user facing search (e-commerce, website) in 3 steps:
Search in action:
- Popular / recent suggestions are based on realtime user analytics
- Relevant searches have additional query modifiers behind the scenes
- Longer searches return an answer
Interactive demo of AI Answers full integration with a faceted search experience
Configuration options now include:
- Rendering sources with labels
- Markdown formatting
- Light / dark theming
- Enable / disable follow-up questions
- Allowing end-users to trigger an AI answer on click or when search ends with a question mark
All of this either as a standalone widget or integrated right into the SearchBox.
See live stuff here - https://t.co/sJQ0H0BPOW
Interactive demos of the top 10 applications of the SearchBox 📦 component (form and function) for #react and #vue - from gen AI to querying from different data sources (popular + curated + index) to different styles (suggestions vs pills vs instant search).
Read the blog post for the details:
https://t.co/1QHV5s4AMW
A peek into embedding AI answers into SearchBox:
https://t.co/a2slXDcGzD
SearchBox experience for AI answers gets seriously good, available in the release candidate React and Vue UI kit releases ✨
Powered by OpenAI + your go-to search engine
Introducing seamless AI Answer integration baked into SearchBox:
- Show suggestions from your search index and provide AI answers
- Latency optimization - Answers stream instantly, just like the demo shows ;)
- Available for latest React and Vue ReactiveSearch UI kits
- Understand the impact of AI answers with end-user analytics
- A point-and-click control plane to tune the AI Answer experience for your users
Live demos with code:
- React https://t.co/5hV8XbjGRq
- Vue https://t.co/NUJtafJRRk
What can you build with this:
- Question answering for your documentation or help hub (just like I did!)
- Question answering based on your domain data
- Incorporate frequently asked user questions and answers into searchbox
Build AI Answering for customer support, docs and internal knowledge bases with ReactiveSearch UI kit for React and Vue.
Question answering works with a first pass to Elasticsearch, and the top hits are sent as context for ChatGPT to source its answer from.
You can read the whole post over here: https://t.co/AImdgqT0jX
ReactiveSearch offers a supercharged search experience with an OpenAI integration with Elasticsearch and OpenSearch, and AI Answer UI components for React and Vue
Build an AI Answer interface for your domain using ElasticSearch and ChatGPT with React (or Vue).
ReactiveSearch simplifies integrating Gen AI into a faceted search experience with a native OpenAI integration and offering configurable UI components.
https://t.co/AImdgqT0jX
AIAnswer can be configured with props to allow for follow-up questions, displaying sources like Bing AI does, enabling/disabling user feedback and custom rendering parts or the whole of AIAnswer