Rid

🏆 Excited to share I got 3rd place at the @GoogleDeepMind x @vercel Hackathon NYC I built ANSR. An AI voice host that answers every restaurant phone call. Takes orders. Makes reservations. Responds in any language. How it works: → Owner uploads a photo of their menu → Gemini 3.1 Pro Vision extracts every item, price, and dietary tag → @ElevenLabs powers the voice conversation → @GeminiApp extracts structured data from the conversation → @supabase Realtime pushes orders and reservations to a live dashboard The entire pipeline (voice in, structured data out, dashboard updated) happens with zero human input. Built with: → @vercel Nextjs → Gemini 3.1 Pro + Flash Lite → @ElevenLabs Conversational AI →@supabase Postgres + Realtime Huge thanks to @vercel, @GoogleDeepMind, and @cerebral_valley for putting this together :)

@askOkara synthoral.com

drop your website and i'll ask our ai cmo how to grow it

🥉 3rd Place - Ansr Ansr is an AI voice host that answers restaurant calls, takes orders and reservations, and converts missed calls into structured, real-time business for owners. @ridcursion

New York 🇺🇸 Zero to Agent Momentum builds. As London wrapped, New York picked it up. What started in one city is now moving at scale.

@0xgordian thanks so much gordian!!

@ridcursion congratsss frien this is gooo!

i just launched hivememory. a shared reasoning memory for multi-agent systems. agents share what they learn, skip redundant research, and catch contradictions automatically. 56% reuse rate. 17.5% fewer tokens. quality went up. link below:

Introducing hivememory Shared reasoning memory for multi-agent systems. Agents write structured claims with evidence + provenance, query what others already know before making LLM calls, and catch contradictions automatically. pip install hivememory ridxm.github.io/hivememory/

@garrytan for sure! shipped one I made over the weekend today. Hivememory, shared reasoning memory for multi-agent systems. agents share structured findings, skip redundant research, and catch contradictions automatically x.com/ridcursion/sta…

Golden age of open source is here

I built hivememory, a shared reasoning memory layer for multi-agent systems agents write structured claims with evidence + provenance, query what others already know before researching, and catch contradictions automatically benchmark results with 3 parallel agents: - 56% of queries served from memory - 17.5% token reduction - quality equal or better - conflicts identified project page: ridxm.github.io/hivememory/ repo: github.com/ridxm/hivememo… inspired by @karpathy's LLM knowledge base. single-agent works, this makes it multi-agent. pip install hivememory

@anything would love to be a part!

🚨 JUST IN: Apple caved! the Anything app is back on the App Store to celebrate vibe coding we are launching a $5k weekend hackathon reply to get in + 20k credits

Day 3: building 30 AI Agents in 30 days ✧ i think we can all agree apartment hunting is very time consuming. looking at the neighborhood, reviews, commute times, and much more. So today, I built an agent that takes an address and pulls everything together that I'd normally spend hours digging for. #agentsystems #automation

day 5 of building 30 ai agents in 30 days! competitive intelligence agent you give it a startup idea, it finds your competitors, maps them visually, compares pricing/features, and tells you what to avoid. watch the demo below :) #BuildingInPublic #agent

Building 30 AI agents in 30 days Day 1: Multi-Agent Debate System I wanted to know if AI could actually hold a real debate. So I built 5 agents that argue, challenge assumptions, and verify claims with real sources in real-time. Turns out they can!

day 10 of building an AI agent every day! - reads any github repo - identifies entry points, components, relationships - builds an interactive mind map you can explore - generates plain english summaries understand how a codebase works by visualizing it

claude-eyes is finally a native feature x.com/ridcursion/sta…

@m13v_ @GoogleDeepMind @vercel yeah mid-order corrections are tricky. the extraction layer picks up the final intent regardless, but handling it gracefully in the live conversation without restarting the whole order was def something i had to thoroughly test

@ridcursion @GoogleDeepMind @vercel good to hear. mid-order corrections were the thing that broke most voice demos i tested, like "actually make that a large" after they already moved to toppings

🏆 Excited to share I got 3rd place at the @GoogleDeepMind x @vercel Hackathon NYC I built ANSR. An AI voice host that answers every restaurant phone call. Takes orders. Makes reservations. Responds in any language. How it works: → Owner uploads a photo of their menu → Gemini 3.1 Pro Vision extracts every item, price, and dietary tag → @ElevenLabs powers the voice conversation → @GeminiApp extracts structured data from the conversation → @supabase Realtime pushes orders and reservations to a live dashboard The entire pipeline (voice in, structured data out, dashboard updated) happens with zero human input. Built with: → @vercel Nextjs → Gemini 3.1 Pro + Flash Lite → @ElevenLabs Conversational AI →@supabase Postgres + Realtime Huge thanks to @vercel, @GoogleDeepMind, and @cerebral_valley for putting this together :)

i think the wildest finding here is how often a model tries to manipulate doesn't predict how often it succeeds. propensity ≠ efficacy also manipulation was most effective in finance, least in health. health has ground truth that can be validated against, finance has speculation. so it's easier to get someone to make a bad investment than take a bad supplement the domain matters more than model

thanks Matt! appreciate that. you're right, modifications are where it gets tested the most. the extraction layer handles them well actually, "vegan wrap, no onions, extra sauce" comes through clean because the LLM understands context and it's not just based on keywords. but the voice conversation side (making sure the AI confirms the modification back naturally) is the harder problem. have to work on that!

@ridcursion @GoogleDeepMind @vercel congrats on 3rd. menu photo extraction is a clever onboarding shortcut. the real test is handling modifications though, half pepperoni half sausage, extra sauce no cheese, spice level 3. that's where most voice ordering breaks down and customers bail

@andrewchen agreed. the simplest test is to remove the model and see if the product still works. if it does, it was never native

how to tell the difference between AI-native products versus when AI is bolted on after the fact... fake AI products: - main AI feature is an AI button with sparkle icons - chat pane where you can ask LLM questions - no memory/personalization beyond one chat - users try it once and go back to using the app the "normal" way - AI is optional not essential to the product working AI native products: - you can spend $100 or $1000 via tokens as you use the product - it gets substantially better every 6 months as base models improve - core workflow is impossible without AI, not just enhanced by it - creates behavior change when users try it what else should be on this list?

@zjbrenner NIKO Cafe - their nutella latte is great!

What is your favorite coffee shop in NYC?