
In May 2026 a decentralized venue became the first to run perpetual contracts on individual US stocks using Nasdaq market data, a deal announced on 18 May. The person who spent months negotiating it had, a decade earlier, deferred her Harvard admission by a year to chase a contract with a ballet company. Kaledora Kiernan-Linn is the cofounder and chief executive of Ostium Labs, the Arbitrum-based protocol that lists perpetuals on stocks, indices, commodities and currencies rather than on tokens. She has held that role since the company was founded in 2022, and no change to it has been reported since December 2025.
Her path matters to a crypto trader for a concrete reason. As of Monday 31 August 2026 the Phemex perpetual board carries four tokenized equities, IONQ, MARA, MRK and TEM, and the pricing and data problem those contracts have to solve is the same one Ostium spent two years engineering around. Her record also carries a $20 million Series A, a Forbes list placement, and a July 2026 exploit that drained $23.75 million from the protocol's liquidity vault. All three belong in the same profile, and the third one is where this article corrects a number Phemex itself published.
Why the Nasdaq Deal Is the Right Place to Start
Getting exchange-grade equity data onto a public blockchain is a licensing problem before it is an engineering one. Market data is a revenue line for every major exchange operator, sold under contracts that specify who may see a price, how it may be displayed, and how far it may be redistributed. Pointing that feed at an onchain venue where anyone with a wallet can read it breaks most of the assumptions those contracts were written under, which is why Kiernan-Linn described the deal as taking "months of discussion and downstream engineering effort around data integrity, security, display policies and more to get us here."
The result is the part traders should care about. A perpetual futures contract is only as good as the price it settles against, and equity perps have historically settled against feeds that thin out or go stale when the underlying cash market is closed. A licensed primary feed narrows that gap. Equity perpetuals accounted for close to 20% of real-world-asset perp activity around the time of the announcement, and the flow in that category is the reason a specialist DEX and a centralized board ended up listing the same kind of instrument within a few months of each other.
The Career Nobody Would Have Modeled
The Harvard Crimson profiled Kiernan-Linn on 30 September 2014 under the headline "On Her Toes," and that piece is the most reliable account of her dancing years because it was written while she was living them. She started at the San Francisco Ballet School at six, moved to the School of American Ballet at eleven, and by the autumn of 2014 was dancing in Boston Ballet II, taking her first corps role in Swan Lake that October under the stage name Kaledora Fontana.
One detail deserves flagging rather than smoothing over. Fortune's December 2025 coverage of the Series A places her at a different, European company for four years, and the contemporaneous Crimson profile names Boston Ballet II and mentions no such company anywhere. Both accounts cannot be complete, and the 2014 reporting was done at the time rather than reconstructed a decade later, so this profile follows it.
|
When
|
What
|
|
From age six
|
San Francisco Ballet School, per the 2014 Crimson profile
|
|
From age eleven
|
School of American Ballet in New York
|
|
2014
|
Boston Ballet II, first corps role in Swan Lake that October, dancing as Kaledora Fontana
|
|
Before matriculating
|
Harvard admission deferred one year to secure a company contract
|
|
2022
|
Cofounded Ostium Labs at Harvard with Marco Antonio Ribeiro
|
|
Class of 2025
|
Named to the Forbes 30 Under 30 Finance list
|
|
3 December 2025
|
$20 million Series A led by General Catalyst and Jump Trading's crypto arm
|
Two rows carry caveats worth stating plainly. Her Harvard concentration is given as neuroscience in Forbes' own list entry for Ostium Labs, and that is the only sourcing for it found in this research, so treat it as reported rather than confirmed. The same entry describes a company started from a hacker house in Cambridge on roughly $7.5 million from early backers, a figure broadly consistent with the roughly $8 million Ostium had raised before its Series A.
What Ostium Actually Trades
Ostium runs perpetuals on assets that do not live onchain at all. A trader posts stablecoin collateral and takes leveraged exposure to an equity, a stock index, a commodity such as copper or platinum, or a currency pair, with the position settling against Ostium's own quoting system rather than against a spot pool. That is a different business from the token-focused perp venues most crypto traders know, and it puts the protocol in a category where the hard problems are data licensing and oracle integrity rather than liquidity mining.
The scale is real without being enormous. Ostium showed 2,203.9 BTC of open interest across 44 perpetual pairs on the CoinGecko derivatives snapshot taken on Monday 31 August 2026. The same snapshot lists a 24-hour volume field of 34.44 BTC, which reads as a partial bar rather than a completed session, so open interest is the sturdier of the two figures and the one used here. Cumulative volume passed $50 billion across more than 26,000 traders by the time of the Nasdaq announcement, and the Series A valued the 15-person company at roughly $250 million.
The $23.75 Million Exploit, and the Figure Phemex Got Wrong
Phemex covered this incident under the headline "How the Ostium Oracle Exploit Drained $18 Million From an RWA Perps DEX." That $18 million came from the first security-firm estimate published while the attack was still being triaged, and it is low. Ostium's own post-mortem puts the loss at 23,752,746 USDC, about $23.75 million, taken from the protocol's liquidity provider vault. Third-party accounts published over the following two weeks ranged from $18 million to $24 million, and the $23.75 million figure is the protocol's own count of what left the vault.
The mechanics are the part most coverage skipped, and they are why this exploit belongs in a founder profile rather than only in a DeFi exploit roundup. Beginning at 14:18 UTC on Wednesday 15 July 2026, over a window of roughly five minutes, an attacker submitted price reports that were fabricated but carried valid signatures, opened large positions against those prices and closed them instantly to extract an artificial profit. The protocol's oracle checked that a price came from an authorized signer. It did not check that the price was plausible.
Ostium's investigation traced the access to compromised off-chain infrastructure credentials, and the post-mortem is explicit that it found no evidence of a smart-contract failure, stating the team had "no evidence this incident was a result of a vulnerability in Ostium's smart-contract code logic or a compromise of the multisigs that govern the protocol." Trader collateral was untouched, with user margin remaining inside the trading contracts throughout. Trading was paused within 60 minutes of the first malicious transaction and reopened in stages on 23 July.
That distinction is worth internalizing if you trade anywhere onchain. Audits cover contract logic. They do not cover the servers, signing keys and keeper bots that feed those contracts, and a venue can pass every audit it commissions while its real attack surface sits on infrastructure no auditor was hired to look at. The same lesson has now been paid for by several perp DEXs at very different scales.
How to Weigh a Founder Who Has Already Had Her Worst Day
Most founder profiles get written before anything goes badly wrong, which makes them close to useless as a guide to how the person behaves under pressure. Kiernan-Linn's has already been tested. She confirmed the exploit window publicly, stated that positions were frozen rather than lost, promised a technical post-mortem, and shipped one. Trading resumed eight days after the halt, with a stated recovery plan for liquidity providers still to follow.
The comparison worth making is not with the founders of larger token-native venues, though that path is well documented. It is with the operators of the tokenized equity products now appearing on centralized boards and on Solana. Every one of them faces the same three problems Ostium hit first. Getting licensed data, keeping a price feed honest between sessions, and surviving the day someone gets into the machinery that signs it.
Frequently Asked Questions
Who is Kaledora Kiernan-Linn?
She is the cofounder and chief executive of Ostium Labs, a protocol on Arbitrum offering perpetual contracts on stocks, indices, commodities and currencies. She trained as a classical ballet dancer before attending Harvard, founded Ostium in 2022 with a classmate, and raised a $20 million Series A on 3 December 2025 at a valuation of roughly $250 million.
Was Kaledora Kiernan-Linn really a professional ballet dancer?
Yes, and the contemporaneous record is a September 2014 Harvard Crimson feature placing her in Boston Ballet II with a corps role in Swan Lake that autumn. Forbes describes roughly five years as a professional dancer before her Harvard degree, and later coverage names a different company that the 2014 reporting does not mention, so the specific company history is not settled across sources.
Does Ostium have a token?
No, and as of Monday 31 August 2026 there is no Ostium token to buy, so any listing claiming otherwise should be treated as a decoy until the protocol says so itself. Traders who want exposure to the chain Ostium runs on trade ARB, and traders who want equity perpetuals can find tokenized equity contracts on centralized boards.
Who cofounded Ostium with her?
Marco Antonio Ribeiro, a Harvard classmate and a former competitor in the international physics, biology and chemistry olympiads. Forbes describes the two of them starting the company out of a hacker house in Cambridge, and both were named to its 30 Under 30 Finance list for the work.
Bottom Line
The most useful thing in Kiernan-Linn's record is not the ballet, it is the sequence of the last ten months. A Series A in December, a licensed Nasdaq data deal in May, a $23.75 million infrastructure breach in July, and a restart eight days later with the cause named in public rather than buried. Watch two things from here. The liquidity provider recovery plan, which was still being finalized when the post-mortem shipped and is the clearest test of how the company treats the people who funded its vault, and open interest in the Nasdaq-powered equity perps, which is the number deciding if licensed data actually pulled in the sophisticated flow it was meant to attract. For anyone trading tokenized equities on a centralized board instead, the transferable lesson is cheaper to learn secondhand. Ask what signs the price, not what audits the contract.
This article is for informational purposes only and does not constitute financial or investment advice. Cryptocurrency trading involves substantial risk. Always conduct your own research before making trading decisions.
