Nimble

Announcing Nimble! The web holds the data to help AI take the next leap, but it isn’t a database. So we built a product that makes it behave like one. We’ve raised $75M from @NorwestVP, @databricks, and leading VCs to build a system that enables anyone to create live datasets from the web, instantly queryable by AI Agents.

Built by Tobi Atte, Georgia Danzger, Shay Palachy Affek, and Amaury Desrosiers during Nimble’s internal hackathon! NYC Gluten-Free Finder uses Nimble’s Web Search Agents to turn scattered local restaurant data into a map-based discovery tool.

[#Hackathon] Project 5 of 6: Finding gluten-free spots in NYC should be easier than opening 20 tabs. When you give Nimble Skills to Claude Code, you can build an app that finds these options immediately! More ⤵️

Built by Hagay Knorovich, Guy Arad, May Elbar, Viki Gohman, Ohad Berenstein, Liel Elis Machluf during Nimble’s internal hackathon!

Hackathon project #4: Upload you grocery list, type in your zip code, and find the cheapest place to get each item! When Claude Code has access to Nimble's web search, creating apps like this takes a matter of minutes.

In just a few minutes, Nimble and Claude Code work together to: ✔️ build a search plan to locate relevant jobs ✔️ pull fresh LinkedIn listings ✔️ deduplicate and clean data before sending ✔️ deliver results to Telegram

[#Hackathon] Introducing our third hackathon project: Nimble Telegram Job Scraper Built by Maor Daniel during Nimble’s internal hackathon, this project uses Nimble’s Web Search Agents via the Nimble skill in #claudecode to turn @LinkedIn job hunting into a live @telegram feed 👇

Built by Nimble's Avishai, Aviv Tsadik, Eyal Halfon, and Noam L, Poly Monitor uses Nimble’s Web Search Agents to turn Polymarket data into a live monitor. Bigger idea: builders get leverage when live web data becomes structured, trackable, and actionable.

Imagine if every bet you could place on Polymarket was a searchable database 👀 Sounds hard to build, right? With Nimble and OpenCode, it took minutes. ⬇️

Meet our first hackathon project: Nimble #Clawbot Built by @ilChemla and Roy Arad, it shows how Nimble can turn live Clawbot into a web researching expert! From researching flights to finding the best restaurants, Nimble makes Clawbot more powerful than ever!!! 👇

👉 𝗦𝗲𝗿𝗶𝗲𝘀 𝗹𝗶𝗻𝗲𝘂𝗽: 1. Nimble Clawbot 2. Poly Monitor 3. Nimble Telegram Job Scraper 4. NimbleCart 5. NYC Office: Best Gym 6. NYC Gluten-Free Finder The team with the most likes, comments, and shares will win a prize AND internal Nimble glory. Stay tuned! #Hackathon

6 teams at Nimble built 6 data apps/agents using our platform, showing different ways you can build with Nimble. Over the next four days, we'll be sharing each of the projects and want YOUR help in determining which is the coolest ❄️ More👇

a16z.news/p/your-data-ag… Thanks @JasonSCui and @JenniferHli !!!

Great piece from @a16z on why data agents fail without a context layer. One blind spot: they only talk about *internal* data. Your agents also need context from the web — competitor pricing, market trends, customer sentiment, regulatory changes. The external world is messy, unstructured, and constantly changing. Sound familiar? The same context layer problem exists for web data. And it's just as unsolved. At Nimble, we think the winning AI stack has two context layers: one for what's inside your walls, one for everything outside them. Read more below ⬇️

Impulse buying is too easy these days. That's why one of our users built an AI app that helps find you the best products... Paste any product link → Claude uses Nimble to find cheaper alternatives, analyze reviews, etc to get a buy/skip verdict instantly Here's a live demo with a video doorbell 👇

Building a RAG pipeline? Focus less on prompts. Focus more on data pipelines. Retrieval only works when the underlying data is structured, current, and reliable. This guide walks through the full production architecture. Read more below...

One of the hardest problems in modern data stacks is keeping external data fresh inside your data lake. Web data changes constantly, but building and maintaining pipelines to collect it is expensive and brittle. Upriver + Nimble automate the entire process: - identify the right tables - generate enrichment pipelines - collect live web data - continuously update your Snowflake datasets In this post, we show one example: building a continuously updated distributor pricing table from live web data. Read more below