Riveter’s cover photo
Riveter

Riveter

Software Development

San Francisco, CA 1,170 followers

Data collection by AI agents - no code required. Backed by Y Combinator.

About us

Riveter's AI agents navigate the web and collect data for you. No engineers required. With Riveter, you can quickly integrate powerful AI into your existing workflows. E-commerce companies save hundreds of hours by using Riveter to streamline critical operations like launching new products, writing SEO-optimized descriptions, analyzing pricing and competitor data, reviewing customer feedback, and more. Our founding team brings deep product and engineering expertise from to YC-backed companies Gusto, Retool, and Middesk.

Industry
Software Development
Company size
2-10 employees
Headquarters
San Francisco, CA
Type
Privately Held

Locations

Employees at Riveter

Updates

  • Riveter reposted this

    View profile for Paul K.

    Founding Engineer @ Pica | Product, Code, and Community

    🚀 Yesterday, I came across a Y Combinator post for Riveter. Today, it’s already live as a verified integration in Pica. When it comes to web data for your product, the gap between idea and working solution is enormous. Timelines slip, dev cycles get burned, and before you know it, you’ve stitched together your own stack of scrapers, proxies, and headless browsers. And you’re still knee-deep in maintenance. Riveter flips that on its head. Their all-in-one web search agent replaces your entire scraper infrastructure stack including SERP, scraping, browser infra, proxies, and even PDF/image reading with a single API. So you spend time building your product, not the plumbing. 🎉 Huge congrats on the launch Abby Grills and Cody W. — excited to see what this does for scraping infra.

    • No alternative text description for this image
  • Riveter reposted this

    Riveter is replacing the entire scraper infrastructure stack with a single API. When it comes to web scraping for your product, the gap between idea and working solution is huge. Timelines creep, and engineering time is wasted as teams realize it will take months, and 5+ vendors, to get basic scraping systems running. And then there’s all the maintenance. Riveter’s all-in-one search agent replaces SERP, scrapers, proxy managers, and more. And it knows which tools to use to find exactly the data you’re looking for. It’s even easy for AI agents to use. Agents can configure input and output data within the API request, giving them more access to the internet than ever before. Build faster and make your product smarter with Riveter. Your first 500 searches are free at riveterhq.com Congrats on the launch, Abby Grills and Cody W.! https://lnkd.in/gV9yrPgJ

  • The Riveter API can bring web data into your app in minutes. It's incredibly easy to set up and removes the need for handling scrapers, SERP, proxies, etc. yourself. Here's Co-founder Cody W. quickly vibe coding an app about wines, with all data for the app populated by Riveter!

  • Riveter reposted this

    View profile for Cody W.

    CTO & Co-Founder of Riveter (YC-backed)

    Met with a customer just now who compared Riveter to 10 competitors and got some pretty great quotes: - "I'm just going to put you out there on a pedestal. I think your output is best in class." - "I don't how you're working your magic, but it seems to be working." - "You guys are different." 🤙🤙🤙 If you're sick of building custom web scrapers to get structured data from the web try us out for free! https://lnkd.in/gC_cyADF

  • Riveter reposted this

    View profile for Abby Grills

    CEO of Riveter (YC F24) | Ex-Gusto, Middesk

    Demo Day for Y Combinator’s Summer batch is this Tuesday! Want a ✨ free spreadsheet with all the YC S25 companies? ✨ 💬✨Reply with "S25" and I'll DM you the spreadsheet with details on all companies, what they’re building, details on the founders, and more. Made in minutes with Riveter (YC F24).

    • No alternative text description for this image
  • We're excited to announce that you can sign up for Riveter and run 500 cells for free! Riveter is the most powerful AI search agent on the market. We've made AI search cheap, fast, and accessible to anyone. In minutes you can set up a project, write prompts to add columns and click Run. You'll have a swarm of AI agents navigating the internet to find exactly the data you need in every cell. Our custom browser use tooling can make Google searches, scrape and navigate websites, and even read PDFs and images to find the best answers possible. Along with your results, each cell has sources, reasoning, and an audit log of how the agent found the answer.

    View organization page for Y Combinator

    1,501,667 followers

    Riveter is bringing the power of Perplexity to the scale of the spreadsheet. Their AI agents, custom scrapers, and integrations pack the power of browser use and automation for search, at a cost that allows for scale. Riveter makes it easy: 1. Add a new column with only a prompt 2. AI agents search the web, crawl websites, read PDFs/images, and navigate to find answers 3. Get tabular, consistent formatting across the entire column Most teams still spend hours (or hire UpWorkers/contractors) to manually copy product specs, check competitor pricing, pull custom KYB/KYC data to assess customer risk, or scrape filings. Perplexity and ChatGPT are great for single questions, but they can’t handle large amounts of data at once. And they’re working with a single agent and context window instead of one in every cell. And Riveter’s web browsing can go further—even reading PDFs and images that may contain answers. With Riveter, anyone can automate complex, auditable web research without coding. Riveter is offering a free trial with 500 cells run for free - sign up at riveterhq.com. Congrats on the launch, Abby Grills! 🚀 https://lnkd.in/gmjh3EeF

  • Have you ever asked ChatGPT to fill in a spreadsheet for you? The outcome is very predictable. It fills in part of the sheet, but then returns most of the table completely empty. Part of what makes Riveter so powerful is our ability to get around the context window limitations of LLMs by running each cell of data individually. When you ask ChatGPT to fill in a spreadsheet, it's trying to: - Search the web or use the LLM to predict the answer - Format the answer for the cell - Do this *times* however many cells need answers This doesn't work because ChatGPT runs out of memory, it's context window is surpassed, and it quickly hits the token limit for a given response. Riveter gets around this by: 1. Running an individual Agent in each cell, giving the cell the full context window size 2. Smart selection of information from scrapers and search to maximize the content window Give Riveter a try for free: https://www.riveterhq.com/

    View profile for Abby Grills

    CEO of Riveter (YC F24) | Ex-Gusto, Middesk

    Why can’t ChatGPT’s memory compete with YOURS? Token Talk #6: understanding context windows and why AI has a "memory limit." What's a context window: A fixed limit on how much text an LLM can process at once — think of it as AI's working memory. Everything beyond this limit gets "pushed out" and becomes invisible to the model. How it works: Remember how transformers look at all words simultaneously from Token Talk #2? They can only look at a certain number of words at once. GPT-4 can handle about 32,000 words — sounds like a lot until you hit the wall. Example: Analyzing a 50-page document with ChatGPT: - Pages 1-20: AI remembers everything perfectly - Pages 21-40: AI starts losing details from the beginning - Pages 41-50: AI has no memory of pages 1-10 at all Ask "What did the introduction say?" and it might confidently make something up because it can no longer "see" the introduction. Real life application: If you give ChatGPT American Express’s 202pg 10-K Annual Report and ask it to find a value, it won’t be able to because the document is too big. It can’t read and “remember” the entire document (see screenshot). Why this matters: Context windows explain why AI can't handle truly long documents, loses track in extended conversations, and sometimes gives inconsistent answers when discussions get lengthy. The solution: Technologies like RAG create a vector store that can be searched, and then relevant sections of documents are provided to the LLM. I’ll explain more in future Token Talks. 🔍 Next Token Talk: How AI companies are solving the "memory" problem with retrieval systems and other techniques. Follow along for more byte-sized AI lessons!

    • No alternative text description for this image
    • No alternative text description for this image
  • Riveter reposted this

    View profile for Abby Grills

    CEO of Riveter (YC F24) | Ex-Gusto, Middesk

    PSA: You don't need to build your own scrapers! Here’s what you DON'T need to build/buy when using Riveter: - SERP APIs (Google search integration) - IP proxy infrastructure for CAPTCHA avoidance - Custom web scrapers - AI agents for tool selection and results - PDF extraction tools - Image processing pipelines - Data transformation layers - Headless browser management Riveter does it ALL for you. Input → Basic spreadsheet of data, add prompts for new columns, click “Run” Output → Exactly the data/information you’re looking for, formatted how you requested. Riveter’s AI search agents search the web, view and navigate web pages, and format results. We've seen this pattern in our demo calls constantly. Teams start building scraper workflows in-house, then quickly discover the rabbit hole of edge cases, rate limiting, anti-bot measures, and maintenance overhead. Riveter handles all the messy parts so you can get straight to using clean, structured data in your applications. The best part? You can try Riveter free at riveterhq.com

Similar pages

Browse jobs

Funding

Riveter 1 total round

Last Round

Pre seed

US$ 500.0K

Investors

Y Combinator
See more info on crunchbase