Panel prep · 25 August 2026 · Funders VC

Eight pitches.
Two minutes
of questions.

10X Founders Demo Day v2.0 gives each founder five minutes on stage and two minutes of panel Q&A — split across six judges. In practice that is one question each, maybe two. So this sheet does not list questions; it ranks them: one to lead with, four if there is time. It is a friendly competition, so each one is written to let a founder show you something, not to defend themselves.

Startups
8
Format
5 min + 2 min Q&A
Judges
6
Reward pool
$150K+ in kind
Anchor
Sonic

The checking behind the questions is marked by how it was done. Verified means an independent source confirms it. Self-reported means only the team says so. Unverifiable means it could not be checked from outside — usually a gap in the public record rather than anything wrong. These notes are for you, not for the room: they are where a good question comes from, not a script to read out. Checked 25 Aug 2026 · DefiLlama API, on-chain contract reads, RPC lookups, project sites

Nomadz

Consumer · Solana

A crypto-native travel aggregator: book stays and events, pay in USDC on Solana, earn XP and NFT "Suitcase" loot boxes on top. The pitch claims 100 airlines, 1M+ stays and up to 50% better rates. Founded by Ivan Maltsev and Vlad Poltavets, Ukrainian team based in Dubai. Solana Foundation put money in — reportedly its first cheque into a Ukrainian startup — alongside a Metaplex DAO grant. Won the Solana Mobile hackathon; official partner for Breakpoint 2025. By distribution it is the strongest name on this list: 21.5K followers, posting daily, real events presence.

The search is not theirs. Loading the London results page, every hotel image comes from cdn.entravel.com, and each property in their own API response carries a ratestellarId and a giataId. Entravel Group is a Danish white-label hotel-booking infrastructure company; RateStellar is its inventory aggregation and room-mapping product. Entravel sells exactly this to crypto platforms — it markets "non-public, exclusive prices… save up to 60%," which is where the "up to 50%" comes from. It raised $7.5M in August 2026 and says it powers 40+ brands and 2.2M hotels. Nomadz is one of those brands, in a list that includes Kraken and KuCoin. Browser network capture of nomadz.xyz/stays · api.nomadz.xyz/api/v1/bookings/search response · entravel.com · ratestellar.com · Tech.eu, PhocusWire, Dealroom · retrieved 25 Aug 2026

What holds up

Verified

Solana Foundation investment, Metaplex DAO grant, Solana Mobile hackathon win. 21.5K followers, posting daily. The product is live and works — 859 London properties returned for a real date range.

Verified

The 30% discounts are real and shown on the results page. They are also Entravel's CUG rates, which is why prices sit behind "log in to see prices" — supplier contracts forbid showing them publicly.

Not found

"100 airlines." No flights tab in the navigation (Search, Events, Quests, Trips only), no airline mention in any page copy, and /api/v1/flights, /flight, /bookings/flights and /air all return 404. Guessed paths are not proof of absence — but nothing on the surface suggests flights exist.

Sourced, not built

Inventory, rates, discount, images and room mapping all come from Entravel/RateStellar. What Nomadz owns is the front end, the USDC checkout, the XP layer and the audience.

Moving numbers

1M+ stays and 50% on the site, 60% in the pinned post, 2.6M stays in March 2025, 2.8M on a podcast. Against a supplier catalogue of 2.2M.

Unverifiable

Zero business metrics anywhere — no GMV, bookings, users, take rate or revenue. Funding amount undisclosed. Circle, Jupiter, MonkeDAO and Bored Apes are on the partner wall; only Solana Foundation, Metaplex and Solana Mobile could be confirmed.

Lead with this

Efficiency

I travel a lot and I used to love traveltech, until I decided it was hard to invest in — thin margins, someone else’s inventory, expensive loyalty. Looking under the hood here I see Entravel and RateStellar, the same stack as forty-odd other brands. So the question I’d genuinely like answered: what in your model creates efficiency they don’t have, and what specifically gets better because this is on-chain?

Open by conceding the category is hard rather than the company. If they answer “audience,” that’s fair — Solana Foundation money and 21.5K followers are real. Then just ask what it converts to.

If there is time

Partner roadmap

Entravel raised $7.5M partly to build stablecoin settlement of their own. How do you see the division of labour with them over the next couple of years — are you going where they won’t?

Flights

I couldn’t find flights in the product yet. Is that the next release, or a deliberate decision to stay with stays for now? Curious which side turned out harder.

The crypto layer

Prices sit behind login because of the closed-user-group rate rules, so that part is the supplier’s. What does the USDC and XP layer actually move — acquisition cost, retention, or an audience that wouldn’t have booked at all?

Your own metric

What’s the one number you look at every week, and how did it move around Breakpoint?

ISLE Finance

RWA lending · Hedera

On-chain trade finance. Creditworthy enterprise buyers issue invoices; ISLE turns those receivables into working capital for the suppliers, funded by stablecoin lenders on Hedera. The framing is the $2.5T global trade finance gap that banks will not serve at viable unit economics. Incubated by BSOS, a Taiwanese supply-chain fintech backed by Taiwan's National Development Fund; Outlier Ventures RWA Base Camp cohort; Hedera Foundation support. The team is the most credentialed on this list — ex-Google, ex-Deutsche Bank, ex-Merrill, ex-Matrixport, all named publicly with full bios. Contracts descend from the Maple v2 architecture.

Ten months of on-chain history, and the whole curve is two deposits. TVL went $681,676 → $929,005 between Oct 2025 and today. Across 306 days there were exactly two day-over-day moves above 3% — 22 Jan (+10.2%) and 6 May (+8.9%). Everything else is interest accruing on a static pool. Net new lender capital in ten months is roughly $143K. DefiLlama protocol API, 307 daily datapoints, retrieved 25 Aug 2026

What holds up

Verified

Live on Hedera mainnet. $929,005 TVL, two USDC pools at 20.59% and 13.56% APY, two published audits, public Solidity contracts.

Verified

Backers: BSOS, Outlier Ventures, Hedera Foundation, Mask Network. Full team page with real, checkable bios.

Unverifiable

Anti Capital and BNB Incubation Alliance — named in the pitch, found in no primary source. Tether as liquidity provider is claimed on the site only.

Scale gap

BSOS advertises a $15M/month financing pipeline. Deployed capital is $929K. That is two orders of magnitude, and the pipeline belongs to the parent's client book, not to ISLE.

Note

Site advertises a flat 12% APY; the pools actually pay 20.59% and 13.56%. Repo activity is thinning — the product contract was last pushed 25 May 2026. ISLE Personal is still "Coming Soon."

Lead with this

The real constraint

Trade finance on-chain is a thesis I like, and I’ve seen a few teams try it — almost all of them found the bottleneck somewhere other than where they expected. Your ten months on-chain look very steady. What’s actually the binding constraint today: sourcing the receivables, or lender appetite? And which one could a room like this help with?

The last sentence turns it into an offer rather than a test. Whichever they name, the other half is the one they’ve just told you is fine — that’s where to follow up gently.

If there is time

Hedera

I’d like to understand the Hedera choice — from outside it looks like collateral composability is limited there, but I may be underrating the ecosystem. What made it the right base?

Track record

How many invoices have been through a full cycle, and has any gone problematic yet? That’s the part I find most interesting — how the underwriting actually behaved.

First loss

How is the first-loss cover sized and funded? Maple learned some expensive lessons on this — curious what you took from them and what you did differently.

The spread

The pools pay 20.6% and 13.6% where trade finance usually clears 8–12%. Where does the spread come from, and what does it look like in steady state?

Kryptos

Infrastructure · API

"Plaid for digital assets" — one API over 5,500+ integrations spanning exchanges, wallets and chains, positioned as the read layer for tokenized finance. What the pitch does not say is that this is a six-year-old company: founded 2020 in Gothenburg as KryptoSkatt, a crypto tax and back-office SaaS, ~40 staff, SOC 2 Type II, 40,000+ users. Kryptos Connect is a pivot layered on top of a running business, not a new company. CEO Sukesh Kumar Tedla chairs the Swedish Blockchain Association and was on the founding team at Telos. Backers include Binance Labs / YZi Labs, Gate.io and Blockchain Founders Fund.

What holds up

Verified

API is GA, not beta. Public docs, TypeScript and Python SDKs, free tier at 10K req/mo, 99.95% uptime SLA. Real developer surface.

Verified

YZi Labs (ex-Binance Labs) and Gate.io are confirmed investors. SOC 2 Type II is real.

Unverifiable

CMC Labs. Named in the pitch, found nowhere on the site or in any aggregator. The co-host list includes Atlas, "backed by CoinMarketCap" — that may be the source of the attribution.

Inconsistent

The integration count is different on every one of their own properties: 5,500+ on the homepage, 5,000+ in the docs and X bio, with the chain and exchange counts swapping places too. The pitch uses the most generous variant.

Thin

~$992K raised across six years against a 40-person team and a claim to be "the rails for all of tokenized finance." No ARR figure disclosed anywhere.

Off-pitch

The founder launched FyniAI on Virtuals in Aug 2025. The FYNI token trades on Base at ~$322K FDV on roughly $590 of daily volume. Not mentioned in the deck.

Lead with this

Two businesses

Six years of real, recurring revenue in crypto is rare and I don’t want to gloss over it. But tax and back-office is a different buyer and a different sales cycle from Connect. How does revenue split between them today, and which is compounding faster? I ask because I’m effectively looking at two companies and I want to know which one I’m being offered.

Framed as respect for the existing business, which it deserves. The split is the whole diligence question and it costs them nothing to answer.

If there is time

Plaid

Plaid now reads crypto exchange accounts through its Investments API. Where’s your edge — depth of CEX connections, the tax computation on top, something else?

Definitions

Your site says 5,500 integrations and the docs say 5,000. What counts as one — a venue, an endpoint, a contract? I’d just like to read the metric correctly.

FyniAI

I noticed FyniAI on Virtuals. How does that relate to Kryptos — separate venture, or part of the same strategy?

Capital efficiency

Under $1M raised across six years with a team of forty is genuinely efficient. Where are burn and runway now?

Kenomic

AI · Token design

Design, validate, launch and manage a token economy end to end. An AI assistant called Keni turns a description into an editable tokenomics model; a simulation engine runs 10,000 Monte Carlo futures over it; then pre-audited contracts, allocations and vesting deploy across Ethereum, Solana, BSC, Arbitrum, Optimism, Stellar and Sui, with treasury, airdrop and buyback tooling after launch. Scored by a proprietary Token Economic Index. Founded by Rayco Tarrida (MIT Sloan) with a seven-person Madrid team. Backed by Heartfelt VC and the Cointelegraph Accelerator. Started life in 2022 as a tokenomics consultancy and is productizing the software it built for itself.

What holds up

Verified

Real company, named team with checkable LinkedIn profiles, coherent product story, Cointelegraph Accelerator confirmed.

Timeline

Their own site says "official platform launch: late 2026." The end-to-end platform in the pitch is not launched. The app root returned an empty shell — no working product could be confirmed.

Self-reported

"1000+ projects supported", "1B+ simulations", "learned from 10,000 token designs." Unreconcilable with a late-2026 launch unless the 1000+ are consultancy engagements from 2022–24.

Unverifiable

Immutable and Fujitsu sit on the logo wall; neither relationship is corroborated anywhere. Yellow Capital also appears — and Yellow co-hosts this event.

Missing

No GitHub, no public docs, no Telegram, no Discord, and the pricing page 404s — while claiming a REST API, an SDK and an MCP server. For an infrastructure pitch that is the weakest signal on the page.

Correct

No token. For a vendor selling tokenomics validation, that is the right answer, not a gap.

Lead with this

Out-of-sample

Token design is an area where I struggle to tell a good model from a well-presented one — a simulation mostly confirms the assumptions you put in. So the thing that would convince me: three tokens you modelled before launch, and how the model looks against what actually happened a year later. Do you have that, even informally?

This is the strongest possible case they could make, so it’s worth asking as an invitation. If it exists, it separates them from a spreadsheet immediately.

If there is time

Services to software

Of the 1000+ projects, how many are the 2022–24 consultancy and how many pay for software today? The transition from services to product is an interesting story in itself.

Contracts

You deploy pre-audited contracts across seven chains — who audits them and how does liability work? That’s a wide surface for a team of seven.

Developer surface

I couldn’t find public docs or a GitHub, though you offer an API, SDK and an MCP server. Is that a deliberate closed beta, or just not got to yet?

Saying no

Does the TEI score ever tell a client no? It strikes me that the willingness to refuse is the entire brand in this category.

A brokerage network for DeFi. Their SDK, Spice Flow, lets any app be used from any chain with one signature, no bridging and no gas on the destination — the user acts with capital they already hold wherever they hold it. Eight chains: Ethereum, Arbitrum, Base, Monad, Polygon, BNB, Robinhood Chain and Citrea. The first open mainnet integration is Reppo, a prediction market for AI training data, where everything is priced in $REPPO and Spice Flow removes the need to acquire it. $3.4M seed led by Hack VC in December 2024, with the Celestia founders, Nick White and Tekin Salimi among the angels. By audience this is the biggest name on the panel — 102.5K followers, posting daily.

Two things the pitch does not mention. First, the $3.4M was raised for a different product — a high-speed app-rollup; the GitHub org is still called pepper-research and still describes Spicenet as "the Unified and Composable Liquidity Layer." The brokerage framing dates to August 2025. Second, mainnet has slipped about two quarters — Q1 2026 target, then Q2, and today it is a public beta with one client, over docs that carry the banner "not recommended for general use yet." Meanwhile TGE is scheduled for Q3 2026, which is this quarter. Spicenet roadmap and pinned posts, Spice Flow docs, GitHub org metadata · retrieved 25 Aug 2026

What holds up

Verified

$3.4M seed, Hack VC lead, December 2024. Hashlock audit with a public report. Eight chains confirmed from their own pinned post. Reppo integration is real.

Verified

102.5K followers, 3,195 posts, daily cadence. Genuinely the strongest distribution here.

Unverifiable

"6k+ wallets used Spice Flow on Reppo." That number appears in no Spicenet primary source. What is sourceable are devnet figures — Brokex 18k, Elitra 11k — self-reported.

Discount it

The points portal has third-party auto-farming bots published on GitHub. Any wallet count from it needs sybil filtering before it means anything.

Audit scope

Hashlock covered pre-mainnet core code. No audit found covering the current Spice Flow EVM / EIP-7702 delegate contracts — the part users actually sign into.

Opaque

48 repos in the org, none of them the Spice Flow SDK or contracts. Integration requests route to a founder's personal Telegram. Not on DefiLlama — no volume or TVL is independently checkable.

Lead with this

Sequencing

Chain abstraction is a direction I believe in, and the audience you’ve built is real. The one thing I’d push on is sequencing: TGE is on your roadmap for this quarter while the product is a public beta with one integration. What does the token unlock on day one that isn’t possible now — and would a second live client before TGE make the launch stronger?

Phrased as a sequencing question rather than a warning. The second clause gives them an easy, constructive answer if they want it.

If there is time

The pivot

The Hack VC round was for an app-rollup; the brokerage network is a 2025 idea. What did you see that turned you? Pivots like that are usually the most substantive part of the story.

The count

Is the 6k+ wallets figure from the points portal? Curious what’s left after sybil filtering, and how many came back for a second transaction.

Economics

Who fronts the gas and the destination capital — you, solvers, or the app? And what does the take rate per action look like without token subsidy?

Signing surface

EIP-7702 delegation gives your contracts authority over a user’s EOA. Hashlock covered the pre-mainnet core — who’s reviewing the delegate contracts?

T3tris

Vaults · Arbitrum

Zero-fee, permissionless vault infrastructure — anyone deploys a vault with no application and no minimum, using async, epoch-settled ERC-4626 so that strategies which sync vaults cannot hold (CEX, RWA, OTC, options) still work. First cohort of the Arbitrum Mentorship Program, where it took a share of the $100,000 demo day pool and grew to 50+ curators. Backed by Arbitrum, Ink and Bpifrance. The $11M in three weeks is real and checkable — it is $13.5M today, and T3tris is the only project on this panel with material third-party capital actually inside its own contracts. The audit list is also serious: Cyfrin, a Lean formal verification by Verity, and an independent researcher.

The $11M is real. It is also two depositors. Reading getGrossTVL() on every counted Arbitrum vault: Gami USDC holds $6.00M and Ellen Capital BTC holds 76.29 WBTC — together about 96% of Arbitrum TVL. It arrived in step-jumps, not growth: $0 on 1 July, $6.02M on the 3rd, $10.88M on the 8th. And of 48 deployed vaults only 9 carry the "verified" flag — a flag set by their own API, which the DefiLlama adapter reads before counting. The dollar amounts are real; the inclusion set is self-curated. Arbitrum RPC direct contract reads · DefiLlama protocol and fees adapters · ecosystem.t3tris.finance · retrieved 25 Aug 2026

What holds up

Verified

$13,480,935 TVL, above the pitched $11M — Arbitrum $12.68M, Robinhood Chain $0.80M, both at all-time highs. Real, on-chain, checkable.

Verified

Genuine security work: a Cyfrin audit with a public report, a Lean formal verification of the performance-fee accounting by Verity, and an independent review. Rare at this size.

Framing

Those three human reviews are listed alongside eleven AI-assisted ones, undated, with the marketing weight on the length of the list. The Verity proof also explicitly excludes oracle correctness, access control and token transfers — which is where vaults actually lose money.

Anonymous

No founder or team member is named anywhere — not the site, not the docs, not the Arbitrum cohort write-ups. $13.5M of third-party capital sits in contracts written by people with no public identity.

Performance

DefiLlama's fees adapter measures total yield generated by the strategies at −$6,116 all-time and −$6,811 over 30 days. The capital is real and it has net lost money over the measured window.

Business model

Zero protocol fee is true; the high-water-mark performance fee goes to curators. T3tris earns on idle capital during async settlement windows, routed to a money market — revenue that rises the longer depositor money sits unsettled.

Note

304 followers and 32 posts ever, against $13.5M. Consistent with a B2B curator play — but it means no independent user reporting exists to cross-check any of this. No source code is public.

Lead with this

The next two curators

$13.5M on-chain in under two months is the strongest verifiable traction on the panel today — congratulations, that’s real. Looking at the composition, two vaults are around 96% of it. Tell me about those curators: how did they find you, and what has to be true for the third and fourth to be that size? And separately, I’d love to know who’s on the team — I couldn’t find names anywhere.

Leading with the congratulation is honest — the number really is the best here. The team question lands softly at the end but still gets asked, and for most funds it’s the gating one.

If there is time

Idle capital

The model earns on idle capital during settlement windows. How do you think about that balance internally — what keeps the windows short?

The verified set

48 vaults deployed, 9 verified. What’s the criterion, and what’s TVL across all 48? I’d like to understand what the DefiLlama number is counting.

Strategy performance

The fees adapter shows net strategy yield slightly negative all-time. Is that the youth of the vaults, or curator selection — how do you read it?

Curator risk

Permissionless means you don’t pick the curators. What catches a bad strategy before settlement — does the NAV oracle do that work?

Mentioned

Prediction markets · Solana

Trade on what people say. Watch a stream, hear a public figure about to drop a word or make a claim, and take YES or NO on it in real time — "mention markets" as a category. Winner of the Colosseum Frontier Hackathon, which drew 2,857 submissions. Built by Taylor Ferran out of Belfast, apparently solo, with a real hackathon record behind him: SSV Network staking work, ERC-4337 accounts, prizes at Encode, Arc and the Solana Privacy hackathon. The markets are real and actively operated — 28 of them, most recently updated two days ago, on Nvidia earnings calls, World Cup fixtures and Fabrizio Romano.

The pitch says the custom LMSR AMM is "deployed to mainnet," settling in USDC. Their own docs say otherwise, in the first line: "Future feature. The mention market AMM is built and deployed on devnet." Querying the published program ID 2oKQaiKx…PdFrU directly: mainnet-beta returns account-not-found; devnet returns an executable program. The Terms also say trading "involves real SOL," not USDC. Caveat worth stating aloud: the docs are five months old and only that one program ID can be checked — a newer, unpublished deployment cannot be ruled out. Solana RPC getAccountInfo on mainnet-beta and devnet · docs pushed 30 Mar 2026 · retrieved 25 Aug 2026

What holds up

Verified

Colosseum Frontier Hackathon winner — named verbatim in the official announcement. Real founder with a checkable build history.

Verified

28 live markets, actively operated, most recent update 23 Aug 2026. Discord has 282 members.

Doesn’t match

"Deployed to mainnet." The published program ID exists on devnet only. Settlement asset is SOL in their own Terms, not USDC.

What is live

The real-money product resells Polymarket liquidity through the Jupiter Prediction API. That is a thin layer on a competitor's book, not their own AMM.

Presentation

The landing page markets are hardcoded mock data — invented events, invented trader counts, a fake leaderboard ending in "you". The real markets page renders empty server-side.

Governance

No legal entity named in Terms, Privacy or Disclaimer. Governing law given as "the laws of the United States." Resolution is fully centralised: admins decide, decisions are final, no oracle, no dispute path.

Context

Not admitted to Colosseum's Accelerator Cohort 5, though most top-25 peers were. "Mention markets" is also already contested — @mentionmarkets operates one, and an analytics vendor for the niche has more followers than its operators.

Lead with this

How it works today

Mention markets is a genuinely fun idea and the Colosseum win says others thought so too. Reading your docs, the LMSR AMM is marked as devnet and the real-money flow appears to run through Polymarket via Jupiter. Walk me through how it’s actually wired right now — and what needs to happen to switch your own AMM on in mainnet?

Asked as “walk me through it,” which lets him correct the record if there’s a newer deployment. Either way you learn what the real product is.

If there is time

Resolution

Deciding whether someone said a thing is a permanent grey area. How do you handle disputes now, and how will that change as size grows?

Subsidy

LMSR needs the market maker to absorb a bounded loss to quote. Who carries that on a market about one word in one podcast, and what does a market cost you?

Jurisdiction

Kalshi got here via a court case. Where does the entity sit and who can legally trade? It usually sets the pace more than the tech does.

Dependency

If Polymarket or Kalshi run mention markets natively — and Kalshi already does some — what stays yours?

Deploi

Private credit · Polygon

Private credit issued as securities that happen to settle on-chain. Senior loan pools become note tokens, distributed to professional investors under a MiFID II wrapper through Assetera, an EU-regulated DLT venue, with ISINs from Nasdaq CSD Riga. Six asset families from consumer credit to litigation finance and NPLs, indicative yields 5.5–17%, a EUR 1bn programme, EUR 10,000 minimum. The issuer is a real, named legal entity — SIA Deploi, Latvia, registration 40203686846 — which is more regulatory scaffolding than anything else on this panel has. Polygon Labs' CEO is quoted in their launch announcement.

Everything institutional about this checks out. Everything on-chain about it does not: no contract address is published anywhere — not on the site, not in the press, not on rwa.xyz — despite live Polygon issuance claimed since May 2026. Every figure on the site (ACV 102.48, 91.2% current, the DPD buckets) is illustrative UI, not live data. And the four headline numbers — $211K revenue, $300M lender volume, $2.8M contracted ARR, $3.5M at $10M pre — have no source anywhere. No funding round has ever been announced. deploi.org, Cointelegraph and Assetera announcements, funding databases · retrieved 25 Aug 2026

What holds up

Verified

SIA Deploi, Latvia, reg. 40203686846, LEI 529900A6ML2X0Q72F756 — an SPV for issuing loan-backed securities. Named team with real LinkedIn profiles. Stockholm and Riga.

Corroborated

May 2026 Polygon launch and Nasdaq CSD relationship, confirmed by Assetera independently and quoted by Polygon Labs' CEO. Note the Cointelegraph piece is a paid placement, not editorial.

Unverifiable

All four pitch metrics. "Lender volume integrated" and "contracted ARR" are also soft framings that can mean considerably less than they sound.

Missing

No contract address, no docs site, no repository, no rwa.xyz listing. For a pitch built on "on-chain settlement and real-time reporting," nothing is externally auditable.

Ahead of itself

The launch release says ISINs were secured. Their own IR page today reads "ISIN: In Progress to be issued" for three series and shows no ISIN string for the fourth.

Correction

The pitch says ERC-3643 pool tokens. The site says ERC-3475 for the note layer, with ERC-3643 as the investor-facing compliance wrapper. Worth getting right before diligence.

Lead with this

Show me one

The regulatory scaffolding here is heavier than anything else on the panel — a Latvian SPV, Nasdaq CSD, MiFID II distribution — and that’s expensive and slow to build, so credit where it’s due. Which is exactly why I’d like to see the on-chain half: could you give me one contract address and one ISIN I can look up myself? Your IR page still shows the ISINs in progress, so I may just be early.

“I may just be early” gives them the graceful exit and still asks the question. If the address exists it takes them ten seconds; if it doesn’t, the answer tells you where the build really is.

If there is time

Collateral

“DeFi-eligible as collateral” sits beside your own note that there’s no active secondary market yet. How do those fit together — has a venue taken these, and at what haircut?

The channel

The pitch reads crypto-native but the product is MiFID II, professional investors, €10k minimum. Which channel is actually pulling right now?

Revenue mix

Could you break the $211K down — origination, servicing, or spread? And what has to happen for the $2.8M contracted to become recognised?

Underwriting

Litigation finance and NPLs are underwriting businesses more than infrastructure. Who underwrites, and where does first loss sit?