Abdelkader Kiari
Senior Protocol Engineer
I design and build the hard parts of onchain systems: matching engines, incentive design, and contracts that hold real money. Took a $110M+ funded protocol from blank repo to mainnet, zero exploits since.
- Onchain matching engines
- Protocol economics & incentives
- Security first engineering
- Gas & performance optimization
Available for the part of a protocol the rest of the system rests on: settlement, matching, and the mechanism design that holds them under adversaries. Remote.
$5M+
USD in OTC settled onchain
15+
contracts on mainnet
142K→28K
gas per trade
14M+
settlements on testnet
Work
I own the contract layer: settlement engine and its economics, blank repo to mainnet
I own Portal to Bitcoin's EVM contract layer and architected the whole onchain system: the atomic settlement engine and the staking, slashing, incentive, liquidity, fee, and accounting machinery around it, from a blank repo to mainnet, zero exploits since.
15+
contracts across two chains
0
exploits since launch
Onchain matching engine: the AMM to orderbook call
Led the call to replace the protocol's AMM with an onchain orderbook, then built the matching engine underneath it: a red black tree over price levels, a linked list per level, and a hash index that reaches any order directly, so cost per operation stops scaling with book depth. 142K→28K gas per trade, and a book that stays fast as it fills.
142K→28K
gas per trade
70-80%
gas reduction across the system
Trustless institutional OTC: block trades that settle across Bitcoin and EVM with no desk holding the funds
I built the institutional OTC settlement product on top of Portal to Bitcoin's atomic core. Two large counterparties agree a block trade off the public book, and it settles trustlessly across Bitcoin and EVM in a single atomic move, no custodian, no bridge, no wrapped assets. Real institutional deals settled on mainnet, on a settlement path I hardened to institutional reliability under load.
$5M+
USD in institutional OTC settled onchain
12% to 0.03%
settlement failure rate under load
Testimonials
A genuine star performer. I was relieved when he stayed.
The best testing discipline I have seen, and he ships reliably every time.
Abdel is a rockstar.
Writing
Curve or Book: The Ordering Rule Decides, Not the Asset Class
Both primitives priced in one unit on the term where they differ, where volatility cancels wherever a maker chooses his own spread and the venue's ordering rule decides whether he can.
Count Every Division Before the Balance Write
A fee is derived, so the design question is where its remainders are allocated. Six call sites on one engine's fill path, three rounding directions, and an enumerated bound on what fragmentation costs.
Experience
Jan 2024 — Jun 2026
Portal to Bitcoin · Senior Solidity & EVM Engineer
Owned the entire EVM contract layer of a noncustodial crosschain settlement protocol, trustless BTC to EVM, from blank repo to mainnet. 15 contracts live since November 2025, zero exploits.
Feb 2022 — Jan 2024
Blockcerts · Senior Solidity Engineer
Solidity engineer with a scope that grew from contract development to owning the company's blockchain architecture and direction.
Jun 2022 — May 2023
Orivium · CTO (part time)
Technical lead for a blockchain gaming project, owning the architecture and the team.
Oct 2021 — Feb 2022
Tok'Art · Cofounder & CTO
Cofounded an NFT marketplace and owned the technical build.
Education
EPITECH · Engineering Degree in Information Technology · 3W Academy · Fullstack Web Development
About
I'm the engineer you hand the part that cannot fail. At Portal to Bitcoin I own the onchain contract layer of a $110M+ funded noncustodial crosschain protocol and I architected it: the settlement engine, the staking and slashing economics, the incentive design that keeps validators and users honest, the accounting and registries under all of it, blank repo to a mainnet launch that has not been exploited since. I own the whole codebase, so I know every piece of it to the floor, and I go as deep on any one of them as the problem needs.
That depth is the point, not the solo part. I go as deep as the hardest problem needs and I don't stop while it's still dangerous. I reversed the protocol's core matching model from an AMM to an onchain orderbook, argued the call on four grounds, then took the matching engine from 142K to 28K gas per trade so the decision stayed affordable to keep. I run external audit as the release gate, not a formality, and hold code under fuzzing and invariants until it ships clean. These are economic systems written in code, and the hard part lives in both at once: the mechanism and the bytecode.
Hand me the hardest onchain component in a system and I'll own it to that same bar, on a team that holds the line as high as I do. On code that moves real money I treat every line, mine or AI drafted, as guilty until review, fuzzing, and external audit clear it.