DeFi, tokenisation,
and DAOs built to
hold up under pressure.
Six blockchain platforms we've designed, built, and deployed - from a $340M DeFi lending protocol to a regulated tokenisation platform for institutional real estate. Real metrics, verified on-chain, no projections.
Metrics from the chain,
not from a pitch deck.
Most blockchain case studies quote TVL and transaction volumes without telling you the contract address. Ours are verifiable - every figure cited in these case studies can be confirmed on the relevant block explorer.
What you'll see here: what the client needed, what we built, what went wrong and how we fixed it, and what the measurable outcome was six months post-launch.
Every protocol in this collection went through internal fuzz testing, external audit (Certik, Trail of Bits, or Sherlock), and a structured remediation process before mainnet deployment. No exceptions.
The economic model - token flows, incentive structures, attack vectors - is specified and reviewed before a single Solidity function is written. Three of the projects below changed their tokenomics significantly during this phase.
Contract addresses, audit reports, and transaction histories are listed for each case study. You can verify every claimed metric independently - we consider that the minimum standard of transparency for this kind of work.
Three platforms, three different
problems, three different solutions.
Each engagement below is described honestly - including the things that didn't go to plan. Real projects have edge cases; case studies that don't acknowledge them aren't telling the whole story.
LendCore - Multi-Collateral Lending Market
A DeFi lending protocol allowing users to borrow stablecoins against ETH, wBTC, and a curated basket of liquid LSTs. Required a novel liquidation mechanism designed to remain solvent during extreme volatility events - a lesson learned from the Aave and Compound stress tests of 2022.
The original brief asked for a fork of Compound V2 with some UI improvements. After the economic design phase, we recommended a ground-up architecture - specifically to handle the risk profile of LST collateral, which has correlated volatility with ETH during depegs. That conversation delayed the project by six weeks. The resulting liquidation mechanism has been stress-tested through three significant market events without a protocol loss.
The oracle integration was the hardest part. Chainlink latency during the May 2023 market event was longer than the liquidation window we'd designed for. We implemented a fallback to Pyth's pull-based oracle within four days of identifying the issue - before it caused a loss. That fix is documented in our post-mortem, which is public.
"The economic design work upfront was uncomfortable - it challenged assumptions we'd been working with for a year. Six months post-launch, it's the reason we haven't had a bad debt event."
RealToken - Institutional Property Tokenisation Platform
A regulated tokenisation platform for institutional real estate - fractional ownership of commercial property assets via ERC-1400 security tokens, with built-in transfer restrictions for KYC/AML compliance, a compliant secondary market, and integration with a UK FCA-registered transfer agent.
The technical challenge here wasn't the smart contracts - it was the compliance architecture. ERC-1400 handles transfer restrictions at the token level, but the KYC data that drives those restrictions lives off-chain with a regulated transfer agent. Building the bridge between the on-chain permission system and the off-chain KYC records without exposing personal data on a public blockchain required a novel architecture using merkle proofs for whitelist verification.
The FCA regulatory perimeter question - specifically whether the tokens constituted "specified investments" under FSMA - took four months of legal review to resolve. We'd scoped for six weeks. That delay was entirely outside our control, but it taught us to front-load regulatory scoping on any project with a compliance dimension. Our current process starts with a two-week regulatory mapping sprint before any technical design.
"The whitelist architecture Sequere designed means we can onboard new investors in minutes rather than days - without any manual intervention from the compliance team. That's changed what the business model looks like."
VaultDAO - Decentralised Protocol Governance System
A governance system for a DeFi protocol with $89M TVL - on-chain proposal lifecycle, token-weighted voting with delegation, a timelock controller with variable delay based on proposal risk tier, and a multi-sig treasury managed by elected stewards. Designed to be resistant to governance attacks, including flash loan voting exploits.
Governance is where DeFi protocols are most vulnerable - not at the contract level but at the incentive level. A large enough token holder (or a flash loan) can pass a malicious proposal if the governance system isn't designed with attack resistance in mind. We implemented a tiered timelock: routine parameter changes have a 24-hour delay, treasury withdrawals require 72 hours, and protocol upgrades require a 7-day delay and a supermajority threshold.
Two governance attacks were attempted in the first year. The first was a flash loan voting attack - caught by our quorum-at-snapshot-block mechanism, which freezes voting power at the block before the proposal, making flash loans useless. The second was a social engineering attempt to pass a malicious proposal through legitimate governance - blocked by the 7-day delay giving the community time to identify and reject it.
"The tiered timelock was the governance engineer's idea, not ours initially. But when the flash loan attack came, it was the reason nothing happened. The 7-day window for upgrades felt frustrating in development; it felt very sensible eight months later."
Six patterns that separate
protocols that hold from ones that don't.
These aren't principles from a whitepaper. They're observations from building, auditing, and watching blockchain protocols under real market conditions.
Every exploit we've studied starts with an economic assumption that didn't hold under adversarial conditions - not a Solidity bug. We spend more time on mechanism design than on implementation, and we've turned down projects where the tokenomics couldn't be made sound.
Three of the six protocols in this collection had oracle-related issues that required post-launch fixes. Not exploits - edge cases in data freshness, latency, and fallback behaviour that only appear under volatile conditions. Chainlink is good. It's not infallible.
Echidna and Foundry invariant tests catch arithmetic edge cases that auditors can miss because they're reviewing code, not executing it millions of times. Manual audits catch logical errors that fuzzing misses because it doesn't understand what the code is supposed to do. You need both.
You can't write a require() statement that stops a large token holder from proposing a malicious parameter change. Governance security comes from timelocks that give communities time to react, transparent proposal reasoning, and quorum mechanisms that prevent low-participation capture.
Proxy patterns let you fix bugs post-deployment. They also mean a compromised upgrade key can drain the protocol. For most DeFi protocols, the right answer is immutable contracts with carefully designed parameter boundaries - not upgradeability. We push back on upgrade requests more than clients initially expect.
Tenderly alerts and Forta detection bots have caught anomalous behaviour in two of our protocols before it caused damage. One was an attempted flash loan attack; one was a misconfiguration in a liquidity pool that would have affected distributions. Neither was in the audit scope because neither existed yet.
Everything here is
verifiable on-chain.
Contract addresses, transaction histories, and audit reports are available for every project. We think that's the minimum standard of transparency for work that handles public funds.
Three more platforms from the same period.
These projects are shorter summaries - same transparency on outcomes, less detail on the engineering.
A production NFT bridge between Solana and Ethereum mainnet, enabling Metaplex NFTs to be locked and equivalent ERC-721 tokens minted on Ethereum - and vice versa. Built for a gaming studio whose assets lived on Solana but whose secondary market was on OpenSea.
A structured token generation event for a Series B fintech - ERC-20 with vesting schedules for team, investors, and ecosystem, a linear unlock mechanism, a DAO treasury allocation, and compliance with OFAC screening on the TGE smart contract.
A consortium blockchain for food provenance tracking - deployed on a private Polygon network with a public anchoring mechanism to Ethereum mainnet for tamper-evidence. Built for a group of four major UK food retailers tracking 340 SKUs from farm to shelf.
Your protocol, designed to
hold up under real conditions.
Book a free 45-minute technical scoping call. We'll review your protocol design, challenge the economic assumptions, and give you an honest view of the risk surface before you commit to development. No obligation, no generic pitch.