Introducing: markprompt.com
An open-source GPT-4 platform for Markdown, Markdoc and MDX with built-in analytics.
00:00
Michael Fester
615 posts
Michael Fester
@michaelfester
- A powerful combo: Effect RPC + TanStack Query. A year ago, during his internship with us, @ethanniser implemented (among a bunch of other amazing things!) an @EffectTS_ wrapper around @tan_stack Query. We haven't looked back since. It's handling queries and mutations in a fully
- Implementing @OpenAI embeddings for docs search in @motifland. It can answer questions from multiple docs sources.
00:00 - Excited to announce that @marieschneegans and I are part of the @ycombinator W24 batch! We’re building @markprompt, AI for customer support. 🌟 ycombinator.com/launches/JtO-m…
- Just shipped: - markprompt.com/docs (built with Motif + Markprompt 🤗) - 5x faster training - Zip, JSON and and raw file uploads - BYO OpenAI API key - Localhost dev - HTML support (thanks for the PR @nickrttn!) Next up: private GitHub repos!
00:00 - Wrote this blog post on how to sync files between browser and disk using Yjs and the File System Access API. It works really well! motif.land/blog/syncing-t…. Excited to see more progress in this space 🚀
GIF - Inspired by @shadcn and @radix_ui, we revamped the @markprompt React library to make it easy to build custom prompt interfaces using headless components, or use a 1-line component with CSS variables. 📘 Docs: markprompt.com/docs#components 📄 Source: github.com/motifland/mark…
00:00 - How we keep the runtime alive when processing streams at @14__ai: using @EffectTS_ Stream + Deferred. Every customer conversation is not just streamed to the user, it is also tracked in parallel so we can power downstream workflows like summarization, triaging, urgency alerts,
- Today, we’re announcing @14__ai, the Cursor for customer support. Customer support is evolving fast. The old model - humans triaging queues in clunky ticketing systems - doesn’t scale. And slapping an AI layer on top of legacy tools doesn’t cut it anymore. We can do so much
00:00 - We made some design optimizations to the markprompt.com landing page: ◆ Cover: @splinetool (thanks @zenorocha for the pointer!) ◆ Cobe (globe) and React Wrap Balancer by @shuding ◆ Path tracing w/ @framer Motion by @raunofreiberg ◆ Icons by lucide.dev
00:00 - Replying to @michaelfesterIt's based on the following amazing stack: - @nextjs - Framework - @vercel - Hosting + edge runtime - @openai - Models - @supabase - DB + auth - @upstash - Redis + rate limiting - @markdoc_dev, @mdx_js - Content - @tailwindcss - CSS - @plainsupport - Support - @stripe - Payments
- More on the DocsGPT integration in @motifland: your trained model is now exposed at an API endpoint, so you can build your own prompt component and include it on your website, e.g. for querying docs and knowledge bases. It provides rich answers, including images and code blocks.
00:00 - Why we decided to stay in San Francisco after the @ycombinator batch? To be close to our users. Nothing beats sitting down by their side and see them in action. Thank you @gabe__nunez for your time, this was very fruitful!
- I am incredibly happy (and relieved!) to share that I’ve just been approved for my Green Card! 🇺🇸