CEO & Founder @ Nimbleway
New York, USA
Joined March 2009
- Tweets618
- Following497
- Followers695
- Likes307
urik retweeted
Web search tells an agent what exists. A browser lets it go there and interact with the page.
But neither tells your agent how to reliably discover and retrieve the exact data it needs.
That’s where expert web context comes in.
Read the blog from our CEO @urielkn on expert web context: nimbleway.com/blog/the-missi…
urik retweeted
It makes sense to optimize model routing at the harness layer instead of the gateway layer if you want to hillclimb on accuracy/cost for any e2e task.
Every task is solved by a combination of a model mixture and agent harness. Every task requires a different mixture of models (+harness logic) to be at the pareto frontier of accuracy and cost.
* If you only optimize the model mixture at the gateway layer, you lose the broader context encoded in the harness and only optimize at the LLM completion layer.
* If you optimize the model mixture at the harness layer, you can make the "optimal" model choices a priori and while in the agent loop
For any given task, the model and harness are probably co-optimized together, so that the model mixture can only exist with this specific harness shape and vice versa
urik retweeted
Introducing Web Search Agents: expert level web search that self-learns your domain to deliver perfect web context, while burning less tokens.
Generic web search tools start from scratch every time they run. Web Search Agents compound their expertise to adapt web search and crawling to your use case — finding the exact data you need, more efficiently.
Learn more below in our blog: nimbleway.com/blog/specializ…
urik retweeted
The Nimble Plugin for Claude is going viral! 820k installs in the last 10 days.
Here's what you get from Nimble that you don't get with Claude's native web fetch:
➡️ Access web content on JS-heavy pages
➡️ Retrieve fresh data instead of cached info
➡️ Build large datasets from the web more efficiently
➡️ Run purpose-built web research workflows
See the difference between Nimble and Claude's web fetch in our benchmark below ⬇️
Build your own vertical search index in 90 minutes: npm i -g
@ nimble-way/nimble-cli Any domain. Any market. The agent searches the live web, compiles a cross-linked knowledge base, and every future search gets context from everything it already knows.
@karpathy @kepano @ObsidianMD
Browse it yourself.
The entire 134-page GLP-1 index is live:
publish.obsidian.md/nimble/
Click any node. Follow any link. Every fact dated and sourced.
This is what vertical search looks like.
When you can trust the data, you can build on it.
From the same 134-page index:
→ Drug comparison tool
→ Investment dashboard
→ Payer coverage lookup
→ Pharmacy price finder
All sourced. All verifiable. All from one domain index.
100 questions. 5 personas:
• Investor analyst
• Pharma strategy exec
• Insurance/payer analyst
• Patient/consumer
• Journalist
Generic search never scored above 18/25.
Vertical search never scored below 19/25.
Won 99 of 100.
How @nimble_search Web Search Agents work:
1. Live web index — real-time data from across the internet
2. Private domain index — 134 compiled wiki pages with entity cross-links
The agent reads your domain knowledge BEFORE searching.
Every query gets smarter because every query has context.
urik retweeted
Here is the link that will take you to the Nimble skill:
skills.sh/?q=nimble
For more information about how this works, check out their documentation: docs.nimbleway.com/
Thanks to the @nimble_data team for collaborating with me on this post.
urik retweeted
Nimble Way raises $47M to give AI agents access to real-time web data techcrunch.com/2026/02/24/ni…
urik retweeted
You can use Claude Code to pull live, structured data from any website.
It's a Claude skill.
This is not about searching the web and returning a wall of text. That sucks.
This skill will return structured data, organized in tables you can work with immediately.
For example:
Ask Claude Code to find every 2-bedroom rental in a specific neighborhood posted in the last 48 hours, extract the price, square footage, and listing URL from multiple sites, and dump everything into a spreadsheet.
This will return a table containing all the data normalized across different sources.
This works using the Nimble skill. It will search the web in real time and return structured output.
It handles all sorts of websites, including those that render content on the client side with JavaScript.
If you build agents that need web data, this skill is a 10/10.
urik retweeted
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.
urik retweeted
Introducing Multiverse: the first AI-generated multiplayer game.
Multiplayer was the missing piece in AI-generated worlds — now it’s here. Players can interact and shape a shared AI-simulated world, in real-time.
Training and research cost < $1.5K. Run it on your own PC.
We @enigma_ai are open-sourcing everything: code, data, weights, architecture, and research.
Here’s a closer look (and how it works!):
urik retweeted
1/ Introducing the Patronus API: powerful AI evaluation models to accelerate your AI development 🚀
- 20% more accurate than ragas on hallucination detection
- Beats Perspective and Llama Guard on safety tasks by 28% and 11%
- Excels in practical domains like finance and customer support
Hundreds of elite AI teams across companies like @hospitable, @ExaAILabs, and Algomo use Patronus to do alpha evals ⚡
Try it out today: app.patronus.ai