Get the Check

We sat down with Christina Melas-Kyriazi, Partner at Bain Capital Ventures. Before VC, she was a product leader at Affirm. She joined when it was a few hundred people and stayed through IPO. Now she backs founders building AI-enabled companies.We talk about:- How her first "scouting" experience wasn't in venture, it was in high school. She recruited a soccer player onto her cross country team who went on to become an Olympian and one of her closest friends for life- What she learned from Max Levchin at Affirm: the best founders hold the highest level of vision and the lowest level of detail at the same time, and that's what allows them to bend reality to their will- Her hot take on why mission-driven companies will out perform; she put $15M into MagicSchool AI and counts Carrot Fertility as one of her most memorable angel checks- Her answer to the question everyone's asking: where will the labs play vs. AI application companies- Why she gives her friends money with zero questions asked if they're starting a company- The positive imposter syndrome loopIf you're a founder exploring the future of AI and services, Christina is only one cold DM away

In this episode of Get the Check, Maya, Anika, and Priya are back with a packed rundown: Travis Kalanick’s new company Atoms, the escalating Anthropic vs DoW fight, and why wearables are suddenly everywhere.
They start with Kalanick’s post Uber arc starting City Storage Systems (including CloudKitchens) in complete stealth. They unpack what the company is trying to do next after rebranding to Atoms. Atoms will use specialized robots to automate core industries like food preparation plus delivery, mining, and transport. They also revisit the Uber drama that led to Kalanick being pushed out, including a DOJ investigation, $245M lawsuit with Waymo, and accusations of a toxic culture that promoted sexual harassment.
Then they shift to the escalating tension between Anthropic and the U.S. government, a story that is quickly turning into one of the defining AI policy battles. What begins as a major defense contract relationship unravels after Anthropic draws clear lines around how its models can be used, rejecting mass surveillance and autonomous weapons. That stance sparks backlash from the Pentagon and the Trump administration, including efforts to label the company a supply chain risk and cut off government use. You can trade on if the Pentagon will designate Anthropic a supply chain risk here, using our Kalshi referral code, which will give you $10 once you trade.
The episode ends with a debrief on wearables: Pebble’s $75 AI ring, Sandbar’s $23M raise and note-taking stack, Taya’s pendant made by ex-Apple employees. The hosts break down why many of these products still feel like they don’t have clear demand, which is why they’re bullish on Fort. Instead of another generic note taking wearable, Fort is targeting a clear use case by building a wearable specifically for strength training and sleep tracking. Speaking of health and wellness, Lotus Health is an AI doctor you can chat with for free anytime. They just raised $41M, and you can download the app today. The pod tried it, and with just our basic info it pulled up every medication we’ve been on in the last 25 years, sent us relevant studies to our demo, and the app can even refer you to real clinicians when needed.
00:00 Maya might be a pro skier01:04 Intro04:50 Why Travis was fired from Uber13:42 The thesis behind Atoms17:11 Hardware as the next frontier19:58 The unexpected reason Maya thinks Atoms won’t work22:57 Anthropic vs Department of War timeline29:45 Anthropic's red lines34:02 Should Anthropic have red lines44:26 Why LLMs change what’s possible in surveillance45:16 Supply chain risk designation50:10 Kalshi odds of Anthropic as a supply chain risk54:27 An update on wearables

This week Maya and Priya sit down with Ishan Gupta, CTO and co-founder of Juicebox, who comes on the pod to announce that Juicebox just closed their Series B!!
Juicebox is the AI recruiting platform letting you find your next hire by just describing them in plain English. PeopleGPT searches 800M+ profiles and surfaces the best fits. No boolean filters required.
Ishan grew up in India watching his dad run a business and always knew he wanted to be an entrepreneur, but had no idea he would drop out of college just three months in to build Juicebox.
Before Juicebox, there were two failed pivots. Tune in to hear why they built a music company and marketplace first, and most importantly why they ultimately decided to build Juicebox.
Ishan breaks down why the hiring manager always knows exactly who they want but never has time to find them, and the recruiter has time but less context.
LLMs change the game because they are the first technology that can actually read a resume the way a human would, picking up on someone’s trajectory, company stage, promotion pace, even GitHub contributions.
Ishan talks about
Finding message market fit and then product market fit
Why recruiting is a zero sum game
If the recruiter / sourcer role disappears
If a human or AI is more biased
Juicebox is hiring across engineering, product, design, sales, and more. The company is high ownership and still operates like a seed stage startup. juicebox.ai/careers.
Now…a word from our sponsors :)
Lotus Health is an AI doctor you can chat with for free anytime. They just raised $41M, and you can download the app today. The pod tried it, and with just our basic info it pulled up every medication we’ve been on in the last 25 years, sent us relevant studies to our demo, and the app can even refer you to real clinicians when needed.
Here’s $10 on Kalshi when you trade $10. Kalshi was one of the first believers in the pod, and they let you trade on anything. No seriously anything, you can trade on sports, elections, and even what Taylor Swift will play first at her next concert. DM us on IG if you want to know what markets we are in. Important disclaimer: This is not financial advice. The information shared is for informational and entertainment purposes only and should not be considered investment, financial, or trading advice. Always do your own research and consult with a qualified financial professional before making any financial decisions.
00:00 Series B Fundraising
03:36 Unexpected reason SF is the best city to live in
07:02 Co-founder meet cute <3
09:38 The two ideas that didn’t work
14:32 LinkedIn’s recruiter tool will die
17:43 LLMs change the talent search game
18:37 Juicebox product
26:47 Why recruiting is a zero sum game
28:23 Will recruiters disappear
31:52 Why AI is less biased than humans in recruiting
34:27 How AI impacts recruiting as a founder
38:30 Who Juicebox is hiring

This weekend was the pod’s Season 4 launch party. To celebrate you can give them 5 stars and increase the chance Maya agrees to a Season 5. 
This week the pod starts with with Bitcoin, which is down about 50% from its peak (~120K to ~65K). They walk through why this crash doesn’t have just one culprit. Trump’s tariff uncertainty, Iranian geopolitical tension, and Bitcoin in key ETFs all contributed. On Feb 5, BlackRock’s Bitcoin ETF (IBIT) sold a lot of Bitcoin, seemingly because it had to rebalance alongside a broader US software/tech selloff, which tied Bitcoin even more directly to tech stocks. They also talk about how Bitcoin is failing the “digital gold” test right now (gold is up while BTC is down), even if it’s working as a decentralized currency, it’s not working as a store of value.
Next they unpack the viral “2028 Global Intelligence Crisis” essay by Citrini, which imagines a bearish AI scenario: white-collar automation triggers layoffs, SaaS gets copied and commoditized, and consumer spending weakens because workers don’t have money to spend post layoffs. The pod has hot takes on where Citrini’s assumptions are underbaked.
They close with World Labs, a startup that just raised $1B to build “world models” for spatial intelligence aka AI that actually understands 3D environments. Co-founder Fei-Fei Li is considered the God Mother of AI, since she was one of the first to advocate for data as a way to improve ML models instead of algorithmic improvement. Early use cases include robotics training and game development. Tune in to hear if World Labs would “Get the Check”.

00:00 Season 4 launch party recap
03:54 Why Bitcoin crashes by almost 50%
14:57 Citrini essay hot takes
23:31 The permanent underclass
35:47 Godmother of AI Fei-Fei Li's World Labs raises $1B

It’s another 6AM morning recording and Anika nods off during the recording - just kidding.
The hosts run first discuss Matt Schumer’s viral essay “Something Big Is Happening” (82M+ views on X), which argues AI labs intentionally made coding the focus, that most people don’t realize how advanced models are, and that we’re 1–2 years from AI autonomously building the next generation. They contrast it with Will Mandis’ response essay “Tool Shaped Objects,” arguing LLMs look like powerful tools but headlines focus on spend (GPU clusters, CapEx) more than real output, likening today’s prompting to Farmville-style busywork.
Next, they cover Anthropic’s new funding and model progress: Anthropic raises $30B (from NVIDIA, Microsoft, GIC, among others), bringing total capital raised to $57B and valuing the company at $380B after starting in 2021. They break down Opus 4.6 launching three months after 4.5, with a 4x context window increase and a big “needle-in-a-haystack” retrieval benchmark jump (18% to 76%), which they frame as improving the ability to find issues in large codebases even if core problem-solving hasn’t changed as much.
Finally, they move to AI in finance and AI insurance. Rogo raises a $75M Series C led by Sequoia, and Meridian raises a $17M seed led by a16z; both target finance workflows like deal flow automation, diligence, market research, modeling, and deck/spreadsheet automation. The hosts argue adoption is still early and research-heavy, and debate whether Meridian’s “Cursor for finance” IDE approach makes sense versus keeping Excel as the output layer, citing Microsoft adding Claude models to Copilot in Excel and a view that entrenched formats (Excel, GitHub PRs) won’t change quickly.
They close with AIUC (Artificial Intelligence Underwriting Company), which raised a $15M seed from Nat Friedman (ex-GitHub CEO) and is cofounded by Ro Kitz (ex-Anthropic product, Center for AI Safety board member) and Rajiv Dani (ex-McKinsey partner). AIUC aims to underwrite and standardize risk for AI agents via insurance standards and audits (including red teaming), addressing liabilities like slurs, hacked agents, and data leaks/violations.
00:00 Intro01:46 ‘Something Big Is Happening’: Essay by Matt Schumer14:45 Anthropic’s $380B Valuation + Opus 4.6 Deep Dive28:02 Will AI replace Excel? Meridian’s ‘finance IDE’ debate38:33 AIUC: underwriting AI risk with standards, audits & insurance