The House Edge

Appearance

Chapter 6 — Research Note (Phase 1) ​

Status: RESEARCH (Phase 1 — written by agent, awaiting Nick's review) Date: 2026-05-14 Working chapter: 06 — The Infrastructure Layer Carries forward from: SPEC.md

How to read this note ​

Four things to know up front:

  1. The chapter has 15+ named firms to develop. This is more named institutions than any prior chapter. The research below covers six anchor firms (Helius, Jito Labs, Flashbots, Titan, BuilderNet, bloXroute) in depth and 14 additional firms in brief, plus confirmations on the verify-existence firms named in the OUTLINE (Astralane, Nozomi, Eden).

  2. JIP-24 passed in 2025 and the Jito fee structure has changed. Pre-JIP-24, the 6% take on Block Engine tips was split 3%/3% between Jito Labs and the Jito DAO. Post-JIP-24, the entire 6% on both Block Engine and BAM fees routes to the Jito DAO treasury in perpetuity — Jito Labs the company no longer takes a direct cut. This is a 2025 change that Chapter 3's footnote 8 framing (correctly describing the Q2 2025 split) does not reflect for the current 2026 state. I'll flag this for a back-edit consideration in Chs 3 and 5.

  3. Eden Network wound down on 12 August 2025. All products shut: Eden RPC, Eden Rocket, Eden Bundles, 0xProtect, Mempool Stream, Tx Explain. The treasury (2,000 ETH) was distributed to EDEN holders through 30 September 2025. The firm cited the "fiercely competitive and costly" MEV relay/builder space. Eden is not load-bearing in 2026 and the chapter will mention it only in passing as an example of the firms that did not survive the consolidation.

  4. Ben Coverston is CEO of both Temporal AND Harmonic. Same individual, two firms — both Solana infrastructure plays operating at different layers. This is significant because it means the Solana infrastructure layer is more concentrated by individual operator than even the OUTLINE suggested.

1. Key claims ​

Each numbered claim is something the chapter is allowed to state. Sources cited inline.

The chapter's structural argument

  1. On-chain trading market structure is determined less by protocol rules than by a small set of infrastructure firms whose business it is to route, sequence, build, and refund trading flow. The named firms in 2026 operate at distinct layers — RPC providers, block builders, relays, block engines, Order Flow Auction operators — and the layer concentration at most levels is between 25% (relays, somewhat distributed) and over 50% (Ethereum block construction, dominated by Titan).

  2. The aggregate scale of this layer is in the mid-hundreds-of-millions to low-billions of US dollars annual run rate in 2026, with no published single aggregate. The figure must be assembled firm-by-firm. As a directional anchor: Solana's 2025 total network revenue was approximately $1.4 billion, of which approximately $720 million was MEV capture; a meaningful fraction of that flows through the named firms below. (21shares 2026 outlook on Solana's 2025 Report Card; referenced via secondary aggregation)

Helius — the anchor firm on the Solana side

  1. Helius was founded in 2022 by Mert Mumtaz (CEO), Liam Vovk, and Nicolas Pennie. The firm raised a $9.5 million Series A in February 2024 led by Foundation Capital (with Anatoly Yakovenko and Raj Gokal participating personally) and a $21.75 million Series B in September 2024 co-led by Haun Ventures and Founders Fund. PitchBook lists total raised at $67.6 million. (Helius — Funding Announcement)

  2. Helius operates at six layers of the Solana stack simultaneously: as an RPC provider (paid tiers $49–$999 per month plus enterprise); as a streaming-data provider (LaserStream, included in Business tier from 7 April 2026); as a transaction-submission service (Sender, free but routing through Helius's paid validator stake); as the network's largest single validator (over 15 million SOL staked, with 99.9%+ slot success and 100% rewards passthrough); as an LST operator (hSOL via Sanctum launchpad); and as one of the four named launch operators of Jito's BAM (alongside Triton One, SOL Strategies, and Figment). (Helius — Pricing; Solana Compass — Helius Validator; Helius — Bitwise SOL ETF partnership)

  3. Helius's customer base includes Phantom, Backpack, Solflare, Squads, Bitwise (exclusive Solana ETF staking provider), and most of the other named wallet apps the book has cited. The firm explicitly markets a 50/50 split on backrun-MEV rebates to the wallets that route transactions through it — the wallet captures half of any extracted MEV value as a refund, with Helius keeping the other half. (Helius — Backrun Rebates Documentation)

Jito Labs / Jito Foundation — the anchor firm on the Solana MEV side

  1. Jito Labs, founded in 2021 and led by Lucas Bruder (CEO), raised a $10 million Series A in August 2022 from Multicoin Capital and Framework Ventures. The JTO token launched in December 2023; as of 14 May 2026 the market cap was approximately $231 million at a price of $0.49, with 469.88 million tokens circulating against a 1 billion total supply.

  2. Jito Labs' product stack: the Jito-Solana validator client (adopted by approximately 94% of Solana stake by Q2 2025 across all variants); the Block Engine (the bundle-auction surface Chapter 4 named); BAM, the Block Assembly Marketplace (mainnet July 2025); JitoSOL (the largest Solana liquid staking token); the Jito DAO; and the TipRouter aggregation infrastructure.

  3. JIP-24 changed the firm's revenue structure. Pre-JIP-24, the 6% take on Block Engine tips was split 3%/3% between Jito Labs (the company) and the Jito DAO. Post-JIP-24 — passed in 2025 — the entire 6% of both Block Engine and BAM tips routes to the Jito DAO treasury in perpetuity; Jito Labs no longer takes a direct fee from these flows. JitoSOL still takes a 4% management fee on staking rewards across the LST. Jito Labs Q4 2024 revenue was approximately $25 million from Block Engine commissions alone; Q3 2025 was approximately $4.7 million; the projected post-JIP-24 annual run rate to the DAO is $15–50 million. (Jito Forum — JIP-24: Jito DAO Receives All Block Engine Fees and Future BAM Fees)

  4. The structural point Jito Labs proved: an MEV middleware company on Solana can capture nine-figure annual revenue and successfully convert that into a DAO/token structure. The firm's products span the full Solana MEV stack — Block Engine processes the trades, BAM rebuilds the blocks, JitoSOL monetises the resulting yield.

Flashbots — the anchor firm on the Ethereum side

  1. Flashbots was founded in 2020 by Phil Daian, Stephane Gosselin, Alex Obadia, Scott Bigelow, and Tina Zhen. Headcount in March 2026 was approximately 55. The firm's strategic direction is led by Hasu. Stephane Gosselin departed over censorship disagreements and now leads Frontier Research / Primev; Phil Daian remains.

  2. Flashbots' 2026 product stack is structurally different from its 2022–2024 stack. The flagship products: Flashbots Protect (2.1 million+ Ethereum accounts; $43 billion+ in cumulative shielded DEX volume; 313 ETH cumulative refunds returned to users); MEV-Share (the selective-disclosure orderflow auction protocol); Flashbots Relay (now approximately 2.15% of payload share on relayscan.io, down sharply from historical dominance); and BuilderNet contribution (Flashbots co-operates BuilderNet as a founding member, having stopped operating a centralised builder in December 2024). (Flashbots — Protect Documentation; BuilderNet — Introducing BuilderNet)

  3. SUAVE — Flashbots' ambitious cross-chain MEV-as-a-service blockchain — was archived in May 2025, with the firm's strategy consolidating around BuilderNet and Flashnet (the network-anonymized mempool design Chapter 3 referenced). The pivot was not an abandonment of cross-chain ambition but a refocus from "build our own chain" to "build the layer that runs across other chains."

Titan / Gattaca — the firm that built half the Ethereum stack

  1. Titan Builder is operated by Gattaca, a London-based crypto trading firm that evolved into block-building infrastructure. Kubi Mensah is CEO and co-founder of Gattaca. No public seed or Series A funding round has been disclosed for the Gattaca/Titan entity (a separately-named "Titan" Solana meta-DEX aggregator raised $7M from Galaxy Ventures in September 2025 — different firm, same name). (Scraping Bits podcast — Kubi Mensah: Titan's Takeover)

  2. Titan's product stack in 2026: Titan Builder (block construction); Titan Relay (titanrelay.xyz); and Gattaca's adoption work on Arbitrum Timeboost (Chapter 3). The single most striking concentration fact in the book: Titan built approximately 51.55% of Ethereum mainnet blocks in the 24-hour window ending 14 May 2026 (relayscan.io live snapshot), while Titan Relay carried approximately 26.65% of relay payload share. One company's builder + relay business is involved in well over half of Ethereum blocks. (relayscan.io — live, May 2026)

  3. Titan's reported builder margin under exclusive flow arrangements (the Banana Gun deal) has reached approximately 17.75%, materially higher than Beaverbuild's pre-BuilderNet ~9% margin. The exclusive-flow business model is itself the seed for Chapter 7. (Observers — How Two Block Builders Monopolized Ethereum Block Production)

BuilderNet — the multi-operator response to Titan

  1. BuilderNet was launched in November 2024 as a multi-operator block-building protocol, co-founded by Flashbots, Beaverbuild, and Nethermind. v1.2 shipped in February 2025; Beaverbuild fully migrated its standalone builder operation into BuilderNet on 6 May 2025. BuilderNet's current share is approximately 24.09% of Ethereum builder market (May 2026 relayscan.io). (BuilderNet — Introducing BuilderNet; BuilderNet — Beaverbuild Migration)

  2. BuilderNet's structural design: each operator runs a node instance; all operators share orderflow equally; an "open-source refund rule" redistributes excess builder profit back to orderflow originators based on contribution to winning blocks. Onboarding is currently permissioned (governed by Flashbots), which has prompted ongoing criticism that the protocol's decentralisation properties are partial. As of May 2026 the founding three operators remain the principal node operators; no major new operator has been publicly added. (EigenPhi guest post — BuilderNet Infrastructure Monoculture)

bloXroute — the multi-chain relay and BDN firm

  1. bloXroute Labs was founded in 2018 by Uri Klarman (CEO, Northwestern PhD), headquartered in Chicago with R&D in Israel. The firm has raised approximately $89.6 million across seven rounds, with a $70 million Series B in April 2022 led by SoftBank Vision Fund 2 (with Dragonfly, Flow Traders, GSR, Jane Street, Lightspeed, and ParaFi participating). (Tracxn — bloXroute profile)

  2. bloXroute operates a multi-chain infrastructure stack: BDN (Blockchain Distribution Network, the firm's core product for fast transaction propagation); private mempool / MEV-defending RPC services; and on Ethereum specifically, two relay variants — bloXroute Max-Profit (neutral, profit-maximizing) and bloXroute Regulated (OFAC-compliant for US-regulated stakers). The combined Ethereum relay share in May 2026 is approximately 25.27% (Max-Profit 13.44% + Regulated 11.83%) — second-largest relay operator after Ultrasound Money. (relayscan.io)

  3. bloXroute is structurally unique among the named firms: it is the only major infrastructure operator that runs a regulated (OFAC-filtered) relay alongside a neutral relay, capturing demand from both regulated stakers (custodians, US-based funds) and pure profit-maximisers in the same product family.

Tier 2 firms — brief institutional treatment

  1. Triton One is a Solana RPC and validator infrastructure firm operating across Solana, Sui, Monad, and Pythnet. Customers include OpenBook, Pyth, and Wormhole; the firm handles hundreds of millions of daily requests. It is one of the four named launch operators of Jito's BAM (with SOL Strategies, Figment, and Helius) and is partnered with the Solana Foundation on the RPC 2.0 initiative. (Triton — RPC 2.0 announcement)

  2. Temporal is the firm DL News reported in 2025 as the team behind the $40 billion-volume Solana "dark" AMM HumidiFi (Chapter 2). Founded and led by Ben Coverston (ex-Citadel Securities, ex-Akuna, ex-MarginFi) with co-founder Kevin Pang (ex-Jump Trading, ex-Paradigm). Temporal also operates the Nozomi transaction-submission service (use.temporal.xyz/nozomi). Portfolio includes investments in DoubleZero, Backpack, and Sanctum. (DL News — Temporal reportedly behind HumidiFi)

  3. Anza was spun out of Solana Labs in January 2024 to maintain the Agave validator client. Leadership: Stephen Akridge (Solana co-founder), Jed Halfon (Chief Strategy Officer), and Amber Christensen (Operations). Funded via Solana Foundation grants; Solana Labs retains a 13% stake. Not a commercial RPC/infrastructure business — Anza is a pure protocol-engineering organisation. (The Block — Anza spinout from Solana Labs)

  4. Jump Crypto is the division of Jump Trading Group that maintains Firedancer (the full C/C++ Solana client rewrite) and Frankendancer (the hybrid client). Frankendancer captured approximately 26% of Solana validator stake within weeks of mainnet launch; full Firedancer launched on mainnet in 2025 and reached approximately 2% of stake by March 2026. Jump Crypto also led Kolibrio's $2M seed (Chapter 3). (The Block — Firedancer mainnet)

  5. Rakurai (developed in Chapter 5) — Ali Rizvi (ex-Apple, ASIC engineering background); $3M seed March 2025 led by Anagram Ventures; closed-source fork of Agave positioned as the "no timing-games" alternative; ~6% of Solana stake by March 2026.

  6. Harmonic (developed in Chapter 3 and 5) — Ben Coverston (CEO; same individual as Temporal) and Jakob Povšič; $6M Paradigm-led seed (late 2025); open block-building aggregation layer for Solana auctioning blocks from Jito, Temporal, JitoBAM, and Paladin. The CEO overlap with Temporal is the chapter's most striking single-operator concentration fact on Solana. (Harmonic — Overview; The Block — Paradigm leads $6M Harmonic seed)

  7. Kolibrio (developed in Chapter 3) — Alex Starikov and Anatolii Padenko; Lisbon-based; $2M seed February 2023 led by Jump Crypto (with Delta Blockchain Fund and Everstake); operates an Order Flow Auction at the RPC layer.

  8. Paladin is a Solana validator-side bot that runs inside the Jito client to prevent sandwiching and extract atomic-arbitrage MEV for the validator. Reward split: 90% to the block leader, 5% to Palidators and stakers, 5% to PAL token holders. Integrated by Chorus One; one of the four named builders that Harmonic aggregates. (Paladin Documentation; Chorus One — Integration with Paladin)

  9. Astralane is a Solana high-speed infrastructure firm operating astralane.io. Live product: the a4.astralane.io indexing platform; Kafka streaming; ultra-low-latency middleware aimed at on-chain market makers, HFT bots, and DeFi protocols. The firm is cited as Helius's HFT-tier competitor for market-maker flow. Funding round size and founders are not publicly disclosed. (CoinDesk — Solana high-speed infrastructure)

  10. Quasar is the third-largest Ethereum block builder at approximately 15.29% share in May 2026. Founders are not publicly disclosed; the rumoured Manifold Trading connection is unconfirmed. The firm is structurally significant — it is the cleanest case of a builder with substantial share but minimal public profile. (relayscan.io)

  11. Aestus is a non-profit MEV-Boost relay run by community contributors (one named: "KuDeTa") from the EthFinance / EthStaker community. Approximately 7.31% of Ethereum relay payload share in May 2026; serves over 650,000 validators as of February 2026. Operates as a public good with no firm-level revenue model. (Aestus — About; Flashbots Collective — Aestus: A Neutral Relay)

  12. Beaverbuild retired its standalone block-building operation on 6 May 2025 and now operates as a BuilderNet operator only. The standalone beaverbuild.org domain still appears with approximately 1.8% residual builder share on relayscan (May 2026), but this is trickle flow — the company's entire team is now dedicated to BuilderNet development. (BuilderNet — Beaverbuild migration)

  13. Eden Network — DEFUNCT. Wound down operations on 12 August 2025. All products shut: Eden RPC, Eden Rocket, Eden Bundles, 0xProtect, Mempool Stream, Tx Explain. Treasury (2,000 ETH) distributed to EDEN token holders at 0.00001506 ETH per EDEN through 30 September 2025. The firm cited the "fiercely competitive and costly" MEV relay/builder space. The chapter mentions Eden only as an example of the firms that did not survive the consolidation. (The Block — Eden Network shuts down)

  14. Special Mechanisms Group (SMG) is a Consensys division focused on blockchain microstructure research, acquired by Consensys (Joseph Lubin's firm) in 2024. SMG acquired MEV Blocker operations from CoW DAO in January 2026. MEV Blocker had served 4.5 million+ users and returned 6,177 ETH in cumulative rebates by the time of acquisition. CoW co-founder Anna George cited SMG's "technical depth and alignment with MEV Blocker's original mission" as the reason for the transfer. The protocol returns approximately 90% of backrun surplus to users and 10% to validators. (Consensys — SMG acquisition announcement; CoW DAO — SMG acquires MEV Blocker RPC)

  15. CoW DAO has processed 73 million+ transactions and hundreds of billions of dollars in protected volume through February 2026. Revenue model: 50% price-improvement share on limit and market orders + 0.1% volume fee on Gnosis Chain / Arbitrum / Base deployments. Solver auction architecture is the canonical intent-based OFA. Launched on Ink (Kraken's L2) on 17 February 2026. (CoW Forum — Monthly Recap February 2026)

Cross-chain take rate

  1. No published single-aggregate measurement of the infrastructure-layer take rate exists. Building bottom-up from named-firm data: Solana 2025 total network revenue was approximately $1.4 billion of which approximately $720 million was MEV capture; Jito Block Engine alone produced approximately $25 million of revenue for Jito Labs in Q4 2024 and approximately $4.7 million in Q3 2025; the projected post-JIP-24 annual run rate to the Jito DAO is $15–50 million; Helius's tiered subscription business and validator stake produce an opaque but substantial number; Ethereum aggregate MEV-Boost payments cumulative since the Merge are approximately $300 million. A directional aggregate estimate for the infrastructure layer's combined take across both chains in 2026 is in the mid-hundreds-of-millions to low-billions of US dollars annual run rate, though this figure has to be assembled rather than read from a single source.

2. Numbers to verify ​

#NumberSourceDateFlag
N1Helius founded 2022; Series A $9.5M Feb 2024 (Foundation Capital); Series B $21.75M Sep 2024 (Haun + Founders Fund); total ~$67.6M raisedHelius funding announcement; PitchBook2022–2024Strong
N2Helius pricing: Free $0 / Developer $49 / Business $499 / Professional $999 (monthly); Enterprise custom; LaserStream in Business+ from 7 Apr 2026 ($100/TB)Helius pricing2026Live
N3Helius >15M SOL staked = largest single Solana validator; 99.9%+ slot success; 100% rewards passthrough; 0% commissionSolana Compass; Helius validator page2026Strong
N4Helius wallet customers: Phantom, Backpack, Solflare, Squads, Bitwise (exclusive SOL ETF), Bitgo, Ledger, Exodus, TrustHelius wallets page; Helius Bitwise ETF post2026Strong
N5Helius backrun-rebate split: 50% to wallet, 50% to HeliusHelius backrun rebates docs2026Strong
N6Jito Labs founded 2021; Lucas Bruder CEO; Series A $10M Aug 2022 (Multicoin/Framework)Multiple aggregator sources2021–2022Strong
N7JTO token launched Dec 2023; market cap ~$231M / $0.49 (14 May 2026); 469.88M circulating / 1B total supplyCoinMarketCap / CoinGecko14 May 2026Live
N8Jito-Solana client adoption ~94% of Solana stake by Q2 2025 (across all variants); JitoBAM ~28% specifically (Mar 2026 per Ch 5)Syndica March 2026Q2 2025 / Mar 2026Strong
N9JIP-24 (passed 2025): 100% of Block Engine + BAM tips route to Jito DAO; Jito Labs no longer takes a direct cutJito Forum — JIP-242025Strong; load-bearing for the chapter's "Jito the firm vs Jito the DAO" framing
N10Jito Labs Q4 2024 revenue ~$25M from Block Engine alone; Q3 2025 ~$4.7M; projected post-JIP-24 annual run rate $15–50M to DAOJito disclosure documents2024–2025Strong
N11JitoSOL: 4% management fee on staking rewards; largest Solana LSTJito docsLiveStrong
N12Flashbots founded 2020 by Phil Daian, Stephane Gosselin, Alex Obadia, Scott Bigelow, Tina Zhen; ~55 headcount Mar 2026Tracxn; Flashbots writings2020 / 2026Strong
N13Flashbots Protect: 2.1M+ Ethereum accounts; $43B+ shielded DEX volume; 313 ETH cumulative refundsFlashbots Protect docs2026Strong; Flashbots-published
N14SUAVE archived May 2025; Flashbots strategy refocused on BuilderNet + FlashnetFlashbots writingsMay 2025Strong
N15Flashbots Relay ~2.15% of payload share (14 May 2026)relayscan.io14 May 2026Live; will drift
N16Titan operated by Gattaca (parent firm); Kubi Mensah CEO; ex-trading-firm background; no public funding round disclosedMultiple sources; Scraping Bits podcast2023–2026Strong on identity; weak on funding
N17Titan ~51.55% of Ethereum builder share (14 May 2026); Titan Relay ~26.65% of relay payload sharerelayscan.io14 May 2026Live
N18Titan reported 17.75% builder margin under exclusive flow (Banana Gun)Observers2025Already in Ch 5; reuse
N19BuilderNet launched Nov 2024 (Flashbots + Beaverbuild + Nethermind); v1.2 Feb 2025; Beaverbuild migrated 6 May 2025; ~24.09% builder share May 2026BuilderNet announcement; Beaverbuild migration2024–2026Strong
N20bloXroute founded 2018; Uri Klarman CEO; ~$89.6M total funding across 7 rounds; Series B $70M Apr 2022 (SoftBank Vision Fund 2 led)Tracxn2018–2022Strong
N21bloXroute combined Ethereum relay share ~25.27% (May 2026): Max-Profit 13.44% + Regulated 11.83%relayscan.io14 May 2026Live
N22Triton One: BAM Node operator + Solana RPC/validator across Solana, Sui, Monad, Pythnet; partnered with Solana Foundation on RPC 2.0Triton RPC 2.0 announcement2026Strong
N23Temporal (Ben Coverston CEO; ex-Citadel/Akuna/MarginFi): reportedly operates HumidiFi; operates Nozomi transaction-submission; portfolio includes DoubleZero, Backpack, SanctumDL News2025–2026Strong on attribution
N24Ben Coverston = CEO of both Temporal AND Harmonic (same individual; two firms)Lightspeed podcast2026Load-bearing concentration fact
N25Anza spun out of Solana Labs Jan 2024; leadership Stephen Akridge, Jed Halfon, Amber Christensen; Solana Labs retains 13% stake; Solana Foundation grant-fundedThe Block2024Strong
N26Paladin: validator-side bot for sandwich prevention + atomic arb; 90/5/5 reward split (leader/Palidators+stakers/PAL holders); integrated by Chorus One; one of Harmonic's four buildersPaladin docs; Chorus One2026Strong
N27Quasar ~15.29% Ethereum builder share (May 2026); founders NOT publicly disclosedrelayscan.ioMay 2026Flag: founder identity unconfirmed
N28Aestus: ~7.31% Ethereum relay payload (May 2026); 650K+ validators served (Feb 2026); non-profit; community-runrelayscan.io; Aestus2026Strong
N29Beaverbuild standalone retired 6 May 2025; residual ~1.8% builder share is trickle flowBuilderNet Beaverbuild postMay 2025Strong
N30Eden Network wound down 12 Aug 2025; 2,000 ETH treasury distributed at 0.00001506 ETH/EDEN through 30 Sep 2025The BlockAug 2025Strong; non-load-bearing in 2026
N31SMG (Consensys division) acquired MEV Blocker from CoW DAO January 2026; MEV Blocker had 4.5M+ users / 6,177 ETH returned at time of acquisitionConsensys SMG announcement; CoW announcementJan 2026Strong
N32Astralane: Solana HFT-tier RPC/data infrastructure; a4.astralane.io + Kafka streaming + ultra-low-latency middleware; active in 2026; founders + funding not publicly disclosedCoinDesk; Astralane2026Strong on existence; weak on internals
N33Solana 2025 total network revenue ~$1.4B; MEV capture ~$720M of that21shares 2026 outlook — referenced via aggregation2025Strong on direction; flag aggregation source

Deliberately not pinned:

  • Quasar's founders and corporate structure.
  • Titan/Gattaca's complete funding history (only Banana Gun margin number is in the public record).
  • Astralane's funding round and founders.
  • The exact aggregate infrastructure-layer take rate across both chains.

3. Contested or evolving claims ​

  • JIP-24 and Jito Labs' revenue model. Pre-2025, Jito Labs took a 3% slice of Block Engine tips (with the DAO taking another 3%). Post-JIP-24 the structure is 0% to Jito Labs / 6% to the Jito DAO. The chapter should describe this transition explicitly because (a) it materially changes how to frame Jito Labs as a firm (the firm now does not capture Block Engine revenue directly), and (b) it suggests a model other infrastructure firms might follow if their tokens have similar DAO structures.

  • SUAVE's archival as a Flashbots strategic pivot. The chapter could frame SUAVE's May 2025 archival as either (a) a failed ambition or (b) a successful refocus around BuilderNet/Flashnet. Both readings are defensible from the public record. My recommendation: frame it as a refocus, with explicit acknowledgement that the firm dropped a major product line.

  • Titan's true competitive position post-ePBS. If ePBS ships (Chapter 5 documented its slip past Glamsterdam Q2 2026), Titan's 51% share could be eroded by a wider competitive surface. The chapter should describe the present concentration without predicting how ePBS will change it. Chapter 12 handles the forward-looking.

  • BuilderNet's permissioning critique. EigenPhi's "Infrastructure Monoculture" post argues BuilderNet's permissioned onboarding undermines its decentralisation claim. The chapter should describe both BuilderNet's stated design (open-source refund rule, multi-operator, TEE attestation) and the critique (Flashbots-controlled onboarding, founding-three-operator dominance).

  • The "Ben Coverston runs two firms" framing. Temporal and Harmonic are structurally distinct businesses (HumidiFi/Nozomi vs an open block-building marketplace) but share a CEO. The chapter should note this but be careful not to imply structural collusion — the firms operate at different layers and don't directly compete.

4. Characters introduced ​

Meet the Infrastructure Provider is the chapter's single "Meet the actor" sidebar — a template-format introduction to the firm type, instantiated against Helius (the cleanest single case spanning multiple layers of the Solana stack).

Named individuals:

  • Mert Mumtaz (Helius CEO; very public X presence)
  • Lucas Bruder (Jito Labs CEO)
  • Phil Daian (Flashbots co-founder)
  • Hasu (Flashbots strategy lead)
  • Stephane Gosselin (Flashbots co-founder, now Frontier Research / Primev)
  • Kubi Mensah (Gattaca CEO / Titan operator)
  • Uri Klarman (bloXroute Labs CEO)
  • Ben Coverston (Temporal AND Harmonic CEO)
  • Kevin Pang (Temporal co-founder)
  • Stephen Akridge (Anza, Solana co-founder)
  • Jed Halfon (Anza CSO)
  • Ali Rizvi (Rakurai CEO, ex-Apple)
  • Jakob Povšič (Harmonic co-founder)
  • Alex Starikov, Anatolii Padenko (Kolibrio co-founders)
  • Anna George (CoW DAO co-founder)
  • Joseph Lubin (Consensys CEO)

Named firms (already covered above in §1 and tabulated in §2).

5. Worked example candidates ​

The SPEC offered three options. The research strengthens the case for one.

Candidate A — Paired Helius / Flashbots firm profiles (the SPEC's recommendation) ​

Two firms profiled in depth as templates for the chapter's argument. Helius spans six layers of the Solana stack (RPC, streaming, transaction submission, validator, LST, BAM Node). Flashbots spans the Ethereum stack differently (Protect at the user-RPC layer, MEV-Share at the OFA layer, BuilderNet contribution at the builder layer, Flashbots Relay at the relay layer). The chapter's structural argument lands: each firm occupies multiple layers; layer occupation IS the business.

  • Pros: maximum content efficiency; both firms have published enough for primary-source material; the comparative move (Solana's Helius vs Ethereum's Flashbots) lets the chapter develop the cross-chain structural argument naturally.
  • Cons: leaves Titan, Jito Labs, and bloXroute as the "other firms" — they get treatment in the subsection prose but not as anchor profiles.

Candidate B — Single-firm deep dive on one firm ​

Either Helius (most spanning) or Jito Labs (most structurally important on Solana) or Flashbots (most influential on Ethereum). The chapter develops one firm as a fully realised case study and uses the other firms as comparison points.

  • Pros: maximum depth on one firm; reader walks away with a complete mental model of how an infrastructure firm operates.
  • Cons: leaves five other anchor firms underdeveloped.

Candidate C — Linear trade-path tracing ​

A trade is followed through the infrastructure stack: wallet → RPC → OFA → builder → relay → validator. Each layer's firm gets a paragraph-long introduction.

  • Pros: structural continuity with Chs 1, 3, 5; the dollar follows the chapter.
  • Cons: the chapter is more naturally a parallel-market description than a sequential trace; the linear approach forces artificial structure.

Agent's recommendation ​

Candidate A — Paired Helius / Flashbots profiles as the anchor, with the other firms (Jito Labs, Titan, BuilderNet, bloXroute) summarised against them in the relevant subsections. This was the SPEC's recommendation; the research confirms both firms have rich enough public material to support the depth.

6. Open questions for Nick ​

Q1 — Word budget. The chapter has 15+ named firms. Even with paired Helius / Flashbots anchor profiles, the prose will run 7,000–8,000 words. Acceptable, or do you want a tighter version with some firms moved to glossary entries only?

Q2 — JIP-24 framing. Should the chapter develop the Jito Labs → Jito DAO transition in detail (it's a structural event), or note it in a paragraph with a footnote? My recommendation: develop it as a paragraph — it's the cleanest case in the book of an infrastructure firm voluntarily handing its revenue stream to a DAO.

Q3 — The Eden Network mention. Wound down 12 August 2025. Mention as a one-sentence "and the firms that did not survive the consolidation" beat, or skip entirely? My recommendation: one sentence. The contrast between active and defunct firms strengthens the structural argument.

Q4 — Ben Coverston / Temporal-Harmonic framing. The CEO overlap is a striking single fact about Solana infrastructure concentration. Should the chapter develop this as a paragraph (it implies one individual operates two structurally important Solana firms simultaneously) or note it in passing? My recommendation: paragraph. It's exactly the kind of "the firms you've never heard of are run by people who run multiple firms" framing the chapter's structural argument needs.

Q5 — Quasar's unknown founders. The chapter cites Quasar's 15.29% Ethereum builder share but cannot publicly attribute the firm to a named operator. Should I (a) note the opacity explicitly ("Quasar's operators are not publicly disclosed"), (b) cite the Manifold Trading connection with a "reportedly" hedge, or (c) skip Quasar entirely? My recommendation: (a). The chapter's structural argument benefits from naming the opacity — a firm running 15% of Ethereum blocks with no publicly identifiable operators is itself the point.

Q6 — Aggregate take rate. The chapter would benefit from a single headline number ("the infrastructure layer captured $X across both chains in 2026"), but no such number is publicly available. Build a bottom-up estimate with explicit hedge, or omit? My recommendation: build the estimate with the methodology shown in the prose; the reader needs some scale anchor.

Q7 — Back-edits to prior chapters. JIP-24 changes the Jito framing in Chs 3 and 5 (the 6% split between Jito Labs and the Jito DAO is now 0% / 6%). The cited Q2 2025 data is still correct for that window, but the current state is different. Should I make small back-edits to those chapters as part of Phase 3 here, or hold for a dedicated back-pass? My recommendation: small back-edits, with a note in REVIEW_NOTES.md.

Sources cited ​

Primary research and protocol docs:

News and aggregators:

Phase 1 is complete. Per the user's compressed-review pattern, Phase 2 (OUTLINE.md) follows immediately.