Every year around this time, someone publishes a “Top 50 Blockchain Leaders” list. Every year, it’s the same faces: exchange CEOs, L1 founders, VCs. People who spend most of their working hours on podcasts, panels, and Twitter Spaces.
This isn’t that list.
The people below are the ones whose code actually runs when you send a transaction. Client maintainers. Protocol researchers. Library authors. Cryptographers. Half of them don’t have Wikipedia pages. Two of them are pseudonyms. The last one on this list just got publicly fired after refusing $5 million on principle — three separate times — and then walked away from the industry entirely.
We started at 25. Then we spent three days fact-checking every entry — dead GitHub profiles, stale blogs, thin verification, generic client-engineer boilerplate. Five didn’t survive the pass, so we cut to 20 rather than pad. A defensible 20 beats a rationalized 25. All of them built things other people are still building on top of. In no particular order — except the last one, who earned the closer.
Kelvin Fichter
OP Labs · Founding Engineer, Technical Lead of EVM Safety
Has been building L2s in the Optimism universe for six years — since before rollups took over from Plasma as the dominant scaling paradigm. Now leads OP Labs’ EVM Safety team: the group that has to be right about consensus changes when everyone else is moving fast. He’s the one you can find in governance forums patiently explaining why cross-chain state reads are harder than they look. Does present at Devcon, but never on a keynote about “the future of onchain economies.” He’ll be on the technical panel three tracks over.
Started as an anonymous blogger writing rigorous essays on Bitcoin security and PoS incentives at a time when most crypto research was pure hype. His work became the intellectual foundation for how the industry now understands MEV — the invisible tax extracted from every Ethereum transaction — and he then went to Flashbots to help architect the system that redistributes it back to validators and users. Still pseudonymous. Still one of the most cited researchers in the space. His Twitter bio doesn’t say “researcher.” It says “chef @ steakhouse.”
If Satoshi built the house, Wuille rewired the plumbing — and then rewired it again. He gave Bitcoin HD wallets (BIP32, the thing that makes “one seed phrase → all your wallets” work), co-authored SegWit, and co-designed Schnorr signatures and Taproot. He’s responsible for more protocol-level Bitcoin code than anyone alive, does almost no interviews, and couldn’t fill a room at a non-technical conference.
transmissions11
Paradigm Research Advisor + MIT · Author of Solmate
Pseudonymous. Minimal public presence. Wrote Solmate — a lean, gas-optimized Solidity library — and essentially handed the DeFi ecosystem a shared foundation it didn’t know it needed. Virtually every serious DeFi protocol built between 2021 and 2024 (Uniswap v3 ecosystem, Seaport, countless forks) is built on top of his code. Left Paradigm in August 2025 to be an MIT undergrad while retaining a research advisor role. The person has never done a conference talk. The code just… runs.
He wrote revm — a Rust implementation of the Ethereum Virtual Machine that has become the de facto standard EVM library across the Rust ecosystem. It’s inside Reth, Foundry, Helios, Optimism, Coinbase infrastructure, Scroll, and virtually every Ethereum zkVM. If you’ve simulated a transaction, forked a chain in a test, or generated a state proof in the past year, there’s a good chance his code did the actual execution.
Merges most of the pull requests into Reth, the modular Rust Ethereum execution client that has been steadily gaining market share since late 2024. Before Reth he was one of the primary maintainers of Foundry — the Rust-based smart contract test framework that ate Hardhat’s lunch. His Twitter is mostly release notes and pointed complaints about typo-fix pull requests from GitHub farming accounts. If Reth ever becomes a majority Ethereum client, half the reason will be that he was reviewing GitHub issues at 3 AM Berlin time for three years.
Formal verification background — his master’s thesis was on proofs of the seL4 microkernel. Moved to blockchain and quietly built out large parts of Lighthouse, one of the two Rust-based Ethereum consensus clients. Author of the tree-states copy-on-write memory model that lets a beacon node hold 128 recent states in the same footprint as four full ones. In April 2023, he co-discovered a novel MEV unbundling attack in discussion with a peer at another consensus client team — the mitigation shipped in Lighthouse v4.1.0. His Twitter bio says he likes “types, specs and proofs.” He is not exaggerating.
The second pseudonymous Solidity-library author on this list, and no, that’s not a coincidence. Solady picked up where Solmate (see #4) left off — even more aggressively gas-optimized, and now the default library set for DeFi contracts written from 2023 onward. Merged into contracts across Manifold, the Uniswap ecosystem, and countless smaller protocols. Ships constantly, communicates almost exclusively via PR descriptions. Zero conference talks. Zero podcast appearances. The library does the talking.
In late 2015 he wrote a GitHub issue proposing a minimal standard for tokens on Ethereum. That issue became ERC-20 — the specification that every fungible token in the world (yours included) is built to. He also wrote the original web3.js library that everyone used to talk to Ethereum before ethers.js came along. Left the Ethereum Foundation to build Lukso — a new L1 focused on Universal Profiles and on-chain identity. Wrote the token standard that everything runs on top of. Most Ethereum users have never heard his name.
Author of the Off-Chain Reporting protocol — the Byzantine-fault-tolerant consensus mechanism that lets Chainlink oracle networks agree on a data point off-chain and post a single signed report on-chain, cutting fees by orders of magnitude. Also of OCR2 and OCR3, which extended the same primitive to power CCIP, Chainlink’s cross-chain messaging. Every major DeFi protocol using a Chainlink price feed is downstream of his work. Former ETH Zürich, IC3 fellow, Google engineer — but you’d only know that from the affiliations line on the papers.
Wrote Hive — the cross-client testing framework every Ethereum execution and consensus client uses to verify they agree on the protocol. Every mainnet upgrade rehearsal, every devnet, every testnet fork readiness check runs through Hive first. He was one of the three people (with Péter Szilágyi — see #20 — and Felix Lange) who refused all three of the Ethereum Foundation’s $5M offers to spin the Geth team out as a private company. Unlike Péter, he still works at the Foundation. Unlike almost anyone else on this list, he’d genuinely rather be writing exploit test cases than being interviewed about them.
Solana has a co-founder who isn’t Anatoly Yakovenko. His name is Stephen Akridge, and his Crunchbase bio literally says he “spends all his time in the codebase” and previously led the GPU backend at Qualcomm and Intel that consistently beat Nvidia. That’s the actual job description. When Solana’s core engineering forked itself into Anza in early 2024 to build the Agave validator client independently of Solana Labs, he was one of the co-founders. You will never see him on a podcast panel about “the future of high-performance chains.” He is the reason the panel is possible.
Ex-D. E. Shaw high-frequency trading engineer who ended up leading the team that built Firedancer — the second independent Solana validator client, written from scratch in C/C++, targeting over a million transactions per second on commodity hardware. Firedancer went live on Solana mainnet on December 12, 2025 after roughly three years of development and over 100 days of continuous testnet operation. By Q2 2026 it was running on more than 20% of active validators. Solana having client diversity at all is largely a function of his team’s willingness to spend those three years replacing the jet engine of an airplane while it was flying. He does keynotes at Breakpoint, but only about scheduling architectures and cache coherency — never about “why Solana wins.”
Was Bitcoin Core lead maintainer for seven years. Signed every release. Reviewed every consensus-critical PR. Never gave a media interview. Burned out and stepped back from the role in 2022 without much ceremony. If you’ve ever downloaded a Bitcoin Core binary, you’ve verified a signature he made. He is the reason a lot of people can trust the code being run under the currency they hold, and he is one of the least famous people on that fact by orders of magnitude.
He and Joseph Poon co-wrote the Lightning Network whitepaper in 2015. Poon moved on and got most of the public credit. Dryja quietly went to MIT DCI and kept building — including inventing Utreexo, a radical Bitcoin UTXO set accumulator scheme that could make full nodes viable on phones. The kind of person who invents two fundamental technologies and still doesn’t have a Wikipedia page in most languages.
Sergio Demian Lerner
Rootstock / FairGate · Rootstock co-founder, Bitcoin security researcher since 2011
Discovered nine vulnerabilities in Bitcoin Core and five in Ethereum’s protocol and clients. Designed Rootstock — the first EVM-compatible Bitcoin sidechain with a working two-way peg — and the merged-mining scheme that lets Bitcoin miners secure it. Also contributed to the design of Ethereum’s Ethash mining algorithm before Ethereum’s move to PoS. No active blog these days, no podcast circuit, no VC hype cycle. Co-founded FairGate in 2023 while remaining a science specialist at Rootstock Labs. His work spans about fifteen years of contributions across the two largest chains, and most crypto readers have never heard of him.
Discovered Halo — the first practical recursive proof composition scheme that doesn’t require a trusted setup — and then wrote Halo2, the productized version. Halo2 is now the proving system underneath Scroll, various zkEVM efforts, Filecoin research, and Ethereum Foundation research. He is currently leading Tachyon, the next Zcash shielded pool upgrade, from Shielded Labs. Cryptographers’ cryptographer. Almost every zk-rollup you’ve heard of has some connection to a paper he co-authored.
Substrate is the framework Polkadot, Kusama, and every parachain (Moonbeam, Astar, Acala, Hydration, dozens more) is built on. Bastian leads the team that maintains the Substrate node itself — the runtime host that all those chains inherit. Sits in a very flat Parity hierarchy where “lead of the Substrate SDK Node team” mostly means reviewing pull requests on Cumulus, consensus, p2p networking, and offchain workers all day. Has been at Parity since 2018 — one of the longest continuous tenures in the space at a single company.
Wrote the Ethereum Name Service — the naming layer that turns 0x4838B106FCe9647Bdf1E7877BF73cE8B0BAD5f97 into vitalik.eth. Every wallet, dapp, and infrastructure provider that supports human-readable Ethereum addresses uses his code, and by extension, so does every user who has ever typed a .eth name. Also author of EIP-165 (interface detection) and EIP-181 (reverse resolution). Left the Ethereum Foundation in 2018 specifically to build ENS as a public good. Ships EthCC keynotes when there’s something to demo, otherwise merges pull requests.
Péter Szilágyi
Independent (ex-Ethereum Foundation) · Geth lead maintainer, 2015–2025
We’re closing with him deliberately. For most of the past decade, Geth ran the majority of Ethereum nodes — software he stewarded. In June 2025 he went public with what happened: three separate offers of $5 million to spin the Geth team out as a private company. He, Felix Lange, and Martin Holst Swende (see #11) refused all three. Then he discovered the Ethereum Foundation had been secretly funding a competing Geth team inside Nethermind — and within 24 hours of raising it internally, he was fired. He hasn’t returned to web3 work since. He’s on this list anyway. Everything running on top of his code will keep running whether he ever comes back or not, and the way he left — three refused $5M offers, then the quiet firing — is arguably the year’s most consequential piece of Ethereum-governance infrastructure.
Sometimes the loudest signal an underdog can send is walking away without a farewell tour.
Categories we couldn’t fit
There are underdog scenes that deserve their own lists. Cosmos IBC has quiet cross-chain protocol architects. The rest of Blockstream (Andrew Poelstra’s Simplicity work, for one) could fill another five entries in Bitcoin research alone. Move-based chains (Aptos, Sui) have compilers and formal verifiers we didn’t reach. Aztec’s PLONK team sits somewhere between “quietly foundational” and “public enough that they’ll show up on some other list next year.” Solo Bitcoin client builders like Davidson Souza (Floresta — the first working Utreexo-based client, funded by Vinteum) are turning Dryja’s research into shipping products right now. Kamil Chodoła, Gary Schulte, Pankaj Garg, and Sam Kozin all made honest cases for inclusion that got cut on tighter grounds of publicly-verifiable work at the time we finalized this.
We had to stop somewhere. Twenty felt like the honest count for what we could defend individually.
What this list is actually about
Twenty people. Six chains. A large fraction of the code that all crypto’s noisier headlines happen on top of.
If there’s a pattern across these profiles, it isn’t that infrastructure people are unusually humble. It’s that shipping and talking are largely mutually exclusive activities. The founders and CEOs who dominate the annual “top 50” lists have institutional support — comms teams, ghostwriters, curated calendars — that the twenty people above largely don’t. Most of them would rather submit a pull request than a CoinDesk quote — which is why when one of them does speak up publicly, like Péter Szilágyi (#20) did about his firing, it registers as newsworthy at all.
The consequence is asymmetric. Attention flows toward visible people. Decisions get made about the infrastructure by people who don’t build it. And when the funding rails wobble — as they visibly are right now for post-EF Ethereum development — the industry discovers, one uncomfortable article at a time, that its plumbing was maintained by people whose names it had barely learned.
We don’t have a clever fix for this. If you build on any of these people’s work, three suggestions with descending cost: know their names (you’re most of the way there if you finished this article); follow their GitHub or X, because release notes are more informative than most crypto news; and if you happen to work at a token project, donate to Protocol Guild — Ethereum core-dev funding is genuinely stretched thin right now, and Protocol Guild is the cleanest pipe to fix it.
The people on this list will not thank you for reading it. Most of them will be mildly annoyed to be on a list at all. That’s the point.