Co:lab Fridays community demos – November recap

Back in October, we relaunched our community meetup, co:lab fridays, with a brand new format. The highlight of our monthly event is now a showcase of the coolest demos created by our community members from around the world. The second wave of demos presented at our November event were just as awesome as the first, and we really enjoyed seeing their creativity and innovative use of the Cohere API. Check out a recap below, or you can watch the full session.

This Month’s Hot Topic: Cohere’s Open Source Strategy

First up was Cohere’s Head of Machine Learning, Ed Grefenstette, who took the mic to discuss how Cohere is shifting to a more transparent and communicative mode of development. We’ll be releasing early versions of endpoints that exemplify new capabilities, applications, and use cases of our LLMs in private and public betas. We’ve also launched Cohere Sandbox, a collection of GitHub repos that include demos and frameworks that are at early or experimental stages of development. Ed highlighted a few Sandbox projects and invited developers to provide feedback, report issues, make contributions, or play around with the Sandbox repo in their own projects.

This Month’s Coolest Demos: Co:ngress, Debait, and Co:here Chat

We then moved on to the focus of Co:lab Fridays: demo presentations. This month, we pitched and discussed three very different applications of generative AI.

Co:ngress by Efkan Goktepe — Efkan built a web application that scrapes transcripts from the Canadian Parliament and summarizes them into easy-to-read blogs using Cohere’s text summarization capabilities. The app makes politics more approachable and helps everyday Canadians to stay informed about what goes on in the House of Commons. See his GitHub repo.

Debait by Vinush Vigneswaran — This conversational AI app is designed to encourage debate around a topic of the user’s choice. Given an article, the user and the AI bot can discuss the article in the form of a debate. The app uses Cohere’s Generate and Classify endpoints. See his GitHub repo.

Co:here Chat by Jonathan Fernandes — Jonathan wanted to improve the efficiency of child-based support systems by automating repetitive parts of the workflow, including summarization, sentiment analysis, log creation, and duration. His app uses Cohere’s Generate, Classify, and Embed endpoints, as well as the sandbox-conversant-lib library currently in development in the Cohere Sandbox. See his GitHub repo.

And the Demo of the Month Is …

After the demos, our co-founder Nick Frosst had the hardest job of the evening: picking his favorite out of the three. He found all the projects amazing in different ways, but in the end he chose Co:here Chat because the project used the spectrum of Cohere capabilities and pieced it together into something that provides tangible value. With this, Jonathan will receive a one-on-one mentoring session with Nick, a special Discord badge, and promotion of his project on our social channels.

We wrapped up by highlighting a few upcoming events, including hackathons, webinars, and AMAs. Co:lab friday community demos happen on the last Friday of every month. Be sure to Register for our next event on February 3, 12 pm ET!

If you have a Cohere-powered demo you’d like to share at co:lab fridays, please fill out our application form.

P.S. Also, join the co:mmunity conversation on Discord.

Co:lab Fridays Community Demos – November Recap