Cohere Coral Hackathon

Join for the opening to be on top with Hackathon rules, Challenge and the Prizes.

Hackathon Opening

Sandra Kublik will join the stream introduce you to Cohere and announce special hackathon mini challenge with Coral!


Cohere Introduction and Special Hackathon Coral Challenge

Join the session with Sebastian Hofstatter

Cohere Introduction. RAG with Cohere

General Q&A (Discord)

General Tech Q&A (Discord)

The workshop with Billy Trend and Harrison Chase

Cohere x Langchain Workshop

Join the office hour on Twitch chat with Cohere team taking the opportunity to consult and improve your project!

Cohere Office Hour

Zain Hasan with be your guide in the world on vector databases and Cohere LLMs. Learn on practice!

How you can search your data using Weaviate with Cohere Models

Hackathon Deadline 

Live Stream: Winners Pitches and Announcement

R2 Rapid Response team building rapid prototyping using various GenAI tools

R2 Rapid Response

dminhk

[Weaviate Blogs QA Chatbot Prototype by R2 Rapid Response Team] 

Providing users simple question answering chatbot interface over Weaviate's blogs posted in 2023 to catch up on vector database and generative AI trends and get relevant information grounded on Weaviate's excellent blog postings in 2023 as a data source.
LangChain template is used for rapid prototyping with LangServe and LangSmith for production ready prototype build.
Hybrid search offers best part of both world by offering a combination of keyword search and semantic search on top of Cohere's Reranker to provide optimum search results and improve user experience.
 Future roadmap includes front end UI for end users with a feedback loops as documented in one of Weaviate's blog posting.

Behind the scenes, we've turbocharged this rapid prototype with Weaviate's hybrid search capabilities. The combination of semantic and keyword search surfaces optimum results from the corpus of blog content. In addition, Cohere's reranker helps picking the best excerpts.  

The rapid prototype was possible thanks to LangChain templates, LangServe, and LangSmith. This allows us to deliver a production-ready user experience on an accelerated timeline. 

Looking ahead, our roadmap includes building out a frontend UI for end users and implementing feedback loops to continuously improve the experience. But even in its current form, this prototype enables anyone to tap into Weaviate's blogs like never before. Just ask a question, and the knowledge is at your fingertips. 

In closing, I'm excited by the potential for this conversational interface to open up new ways to share Weaviate's thought leadership. Thank you for considering our Weaviate Blogs QA Chatbot submission. I welcome any questions you may have. Thank you. 

Rapid prototyping  of Question Answering Chatbot over Weaviate Blogs in 2023 using Weaviate Hybrid Search, Cohere Reranker, and LangChain Template.

"R2 Rapid Response team on Cohere Coral Hackathon Hackathon"

Team Idea

Submission