Vigyata.AI
Is this your channel?

Screensharing Kevin Rose's AI Workflow/New App

30.9K views· 510 likes· 56:25· Feb 2, 2026

🛍️ Products Mentioned (4)

I sit down with Kevin Rose for a live screen share where he walks me through “Nylon,” a personal Techmeme-style news engine he vibe-coded to track AI and tech stories. He breaks down how he pulls from RSS, enriches articles with tools like iFramely, Firecrawl, and Gemini, then generates TLDRs and vector embeddings to cluster stories with real nuance. We dig into his “gravity engine,” an editorial scoring system that ranks stories by impact, novelty, and builder relevance. The bigger theme is simple: with today’s models and workflows, a solo builder can ship wild, high-leverage software fast, then refine by cutting features down to the few that matter. Timestamps: 00:00 – Intro And What Kevin Plans To Demo 03:10 – Techmeme Breakdown And How Signal Gets Ranked 06:44 – RSS Sources, Ingestion, And The Article Pipeline 11:23 – Winner Selection: RSS vs iFramely vs Firecrawl vs Gemini 13:01 – Why iFramely And Firecrawl, Explained 16:37 – TLDRs, Vector Embeddings, And Why They Beat Keyword Search 19:49 – Task Orchestration With trigger.dev And Retries 24:58 – Clusters: Expanding With Search APIs And Discovery 27:07 – The Gravity Engine: Editorial Scoring Rubric 31:31 – Product Management: Gut, Iteration, And Cutting Features 34:53 – Synthetic Audiences And Personal Software 37:03 – What “Success” Looks Like 43:52 – Retention Mechanics And The Idea Browser Example 47:19 – “Blurred Presence” Blog Project From A 12-Year-Old Idea 50:34 – This the best time to build 51:55 – How To Work With Kevin, DIGG Reboot, And VC Today Keypoints * I watch Kevin’s end-to-end pipeline for turning messy RSS links into clean, enriched, clustered stories. * Kevin uses a “winner” judge to pick the best source of truth per field (summary, main content, metadata). * Vector embeddings plus clustering unlock meaning-level grouping that keyword search misses. * trigger.dev gives durable background jobs, retries, and observability for a solo builder workflow. * His “gravity engine” acts like an editorial layer that prioritizes novelty, impact, and builder relevance. Numbered Section Summaries 1. Nylon: A Solo “Techmeme-Level” Build Kevin shows me Nylon, a nights-and-weekends project built to answer a single question: can one person assemble a Techmeme-quality feed tailored to AI velocity. He frames it as personal curiosity first, product second. 2. From Sources To Articles: The Ingestion Spine He pulls from dozens of sources (RSS, Reddit, major tech outlets) and stores everything in Postgres. Each article flows through a status pipeline that tracks enrichment steps and readiness. 3. Enrichment Stack: iFramely, Firecrawl, Gemini Kevin uses iFramely for rich link metadata cards and Firecrawl for deeper crawling, then leans on Gemini as a last-resort “grounded” fill when crawls fail or quality looks weak. A judge picks the “winner” per field so the database keeps the best available representation. 4. TLDRs And Embeddings: Turning Text Into Math He generates a purposely rich TLDR for vector embeddings, stores vectors in Postgres, and uses them to compare meaning across stories. Kevin highlights how embeddings capture nuance like role reversal in similar headlines. 5. Durability With trigger.dev Instead of fragile cron glue, he runs TypeScript tasks as orchestrated jobs with retries, traces, and monitoring hooks. That keeps the pipeline resilient while he develops locally and scales later. 6. Clustering And Expansion: From Three Signals To The Whole Web Once a topic crosses a threshold, he expands coverage via search APIs (Brave, Tavily) to pull in relevant articles outside the RSS set. The cluster page becomes a living dossier with timing, sources, and similarity distances. 7. The Gravity Engine: Editorial Judgment As Code Kevin layers an “editorial vote” system over clusters, scoring dimensions like industry impact, novelty, technical depth, viral potential, and PR-fluff risk. The point is prioritization: a small list of truly worthy items for a specific person. 8. The Meta Lesson: Personal Software And Play As Strategy We zoom out to vibe coding, distribution mechanics, and how “for fun” projects sometimes become the biggest businesses. Kevin shares ways to connect with him through DIG and his Venice studio, plus his view on when capital makes sense. The #1 tool to find startup ideas/trends - https://www.ideabrowser.com/ LCA helps Fortune 500s and fast-growing startups build their future - from Warner Music to Fortnite to Dropbox. We turn 'what if' into reality with AI, apps, and next-gen products https://latecheckout.agency/ The Vibe Marketer - Resources for people into vibe marketing/marketing with AI: https://www.thevibemarketer.com/ FIND ME ON SOCIAL X/Twitter: https://twitter.com/gregisenberg Instagram: https://instagram.com/gregisenberg/ LinkedIn: https://www.linkedin.com/in/gisenberg/ Kevin Rose: x: https://x.com/kevinrose personal website: https://www.kevinrose.com/about Youtube: https://www.youtube.com/@KevinRose

About This Video

In this one, I do a live screenshare with Kevin Rose and we go deep on his personal “Techmeme-style” news engine he vibe-coded called Nylon. The point isn’t “look at this cool app.” It’s watching an end-to-end pipeline: RSS feeds → article ingestion → enrichment (iFramely, Firecrawl, and Gemini as a last-resort ground-truth fetch) → TLDR generation → vector embeddings → clustering. The wild part is how he uses a “winner” judge to pick the best source of truth per field (summary, main content, metadata) so the database stays clean even when RSS is truncated or crawlers get blocked. Then we get into the stuff that actually matters for builders: embeddings + clustering beats keyword search because it groups by meaning, not just matching words. On top of that, Kevin layers a “gravity engine,” basically an editorial scoring rubric that ranks clusters by impact, novelty, and relevance—so you’re not just building an aggregator, you’re building taste into the product. And operationally, he uses trigger.dev to orchestrate durable background jobs with retries and observability, which is exactly how a solo builder ships high-leverage software without it collapsing the second an API times out. The meta takeaway: it’s never been easier to build the messy sandbox fast—your real edge is cutting features down to the few that matter.

Frequently Asked Questions

🎬 More from Greg Isenberg