We are super excited to share that NationGraph has raised $18M USD in our Series A led by @MenloVentures Ventures, with participation from @perplexity_ai, @xyz_vc , @reachfund, and amazing angel investors who have built iconic companies.
Full announcement below 👇
Why does it feel like everything is way over budget and behind schedule nowadays?
It took 21 years and cost $346 Million to build a short bus lane in San Francisco.
The Golden Gate Bridge took 4.
What's changed? Is American Dynamism Dead?
We break it down on Episode 1 of NationGraph News 👇
if you're in waterloo next wednesday come through!
i'll be sharing some more bts stuff in all things startups, growth, venture & scouting for our w27 intern batch at @trynationgraph !!!
How do you take the messiest data in the country and turn it into something a sales team actually wants to open?
That's the problem Jason, our Head of Engineering and his team solve every day.
His team takes the government data we gather, types of data other companies simply do not have, and wraps it into the product our customers use to get ahead of RFPs and win government contracts
"NationGraph to me is a problem that we're trying to solve that has not been solved yet. But the solution is so fascinating that I would put 12, 16 hours some days just into solving it faster.
How do you turn millions of government documents into real buying signals for your team?
That's the problem Mae and her team solve every day.
Mae is our Head of ML and AI. Stanford physics PhD. Former quantitative researcher at Citadel. Mae explains how it works 👇
Her team uses AI and classical ML models to identify relevant signals, filter out the noise, and build systems that get smarter with every customer interaction.
"I think of a signal as some modicum of truth that helps it be more likely for anyone to want to buy your product or engage with you. Any kind of bit of information that helps with that, we call that a signal."
Performance and $$ depend hugely on harness design. In vertical AI, the gains from a great specialized harness >>> gains from more expensive model
inb4 kimi has quirks - @runta can you try with a diff model also
frontierharness.org/
Most public-sector email campaigns start with the same ingredients: a name, title, and email address.
The teams we talk to can usually find the contact. The hard part is knowing what the agency is working on and why a message matters now. 👇️
if you sell to public schools or cities, the hard part isn't finding leads.
it's knowing which of the thousands of government agencies is ready to buy your product now.
we make that research work dead easy for you to do on @trynationgraph.
Last week we launched Compass, the AI agent inside @trynationgraph, built on the deepest state, local and education dataset in America.
So we filmed this to show you what life looks like with Compass.