Partner chain in a Box
Kevin Hammond has spent much of his career working at the point where research becomes usable engineering. His background is academic, but his account of blockchain development is practical. He talks about Haskell, Cardano and partner chains in terms of systems that have to work, stay working and recover when something goes wrong. That is the thread running through his work, from functional programming to Cardano and now to what he describes as a “partner chain in a box”.
Hammond grew up in the south of England, studied at the University of East Anglia in Norwich, completed his PhD there, then moved north through Glasgow and St Andrews, where he became a professor of computer science. Along the way he built a substantial academic record, with 130 published research papers and research grants worth between £20 million and £30 million. His work centred on functional programming and, in particular, Haskell.
“I wrote the first Haskell compiler. I wrote the first parallel Haskell compiler. Helped design what is now the dominant compiler for Haskell, GHC.”
That Haskell background matters because it is central to the engineering culture Hammond brought into blockchain. He says his role in Haskell was to make the language more practical. The original ambition was that Haskell would serve three purposes: research, teaching and real-world computing. In his account, it succeeded in two of the three. It became a strong research language and a practical one, even if it was never the easiest language to teach. That practical side is what later made Cardano a natural fit.
“My role in the Haskell development was to make Haskell more practical.”
When Hammond joined Input Output (IO) in 2019, the attraction was clear. He had spent years investing in Haskell, and here was a major project using it seriously. At IO he moved from academic research into delivery. He led the development team working in Haskell and says his job was less about writing large amounts of code himself than about providing leadership and direction. Before he joined, he says, the team had not yet delivered. After he joined, Cardano delivered six or seven hard forks as well as smart contracts, native tokens, governance and performance improvements.
His most useful point is one that applies far beyond blockchain. Teams often focus on building the first version of a system and underestimate what comes next. Hammond is clear that maintenance is not secondary work. Once a system is live, it has to be updated, released, tested, debugged and improved continuously. In his words, there is “an awful lot of grunt work” needed to keep a system working once the initial delivery stage is over. That is where engineering discipline shows itself.
“There’s a huge difference between doing initial development and having deployed the system.”
That same engineering mindset shapes how Hammond talks about robustness. He argues that Cardano has been built as an engineered system rather than simply a software project. In his account, that means systems designed to withstand shocks and keep operating. He points to Cardano’s record over several years and to one incident in particular, when a transaction caused the chain to fork into two parts. Some nodes followed one chain, others followed another. The important point for Hammond is what happened next. The network recovered and brought itself back together. He calls that self-healing.
He does not present that as perfection. He is clear that engineers are the first to worry about imperfections. But he does present it as evidence of resilience. His description is concrete: a large, complex system, built in more than 100,000 lines of Haskell, hit a serious on-chain problem and recovered. That is the standard he wants to carry into the next phase of work.
That next phase is Ensurable Systems, the company Hammond co-founded with Philipp Kant after leaving IO in June 2025. The aim is to provide custom protocol, ledger and consensus design that builds on Cardano without being limited to Cardano mainnet. The first example is the Tachys protocol, which grew out of work done with Hal8 Apex Fusion and other contributors using the Cardano code base for the Apex Vector and Prime networks. Hammond’s explanation is straightforward. Rather than writing a new code base from scratch, Apex Fusion used the Cardano code base because it was already proven. In doing so, the team found they could generalise the protocol and make it reusable for other partner chains.
The technical argument behind Tachys starts with slot leadership. Hammond says Duncan Coutts identified slots in the Ouroboros Praos protocol that were unused because the leader selection process deliberately chooses only a small fraction of the available slots to mint blocks and doesn’t reveal the slot leaders before blocks are minted. If that level of obscurity is not needed in a partner-chain environment, more slots can be used for producing blocks. Hammond says that alone can produce a fourfold increase in throughput. In a more constrained setting, with fewer validators and a smaller geographic footprint than Cardano mainnet, the gains can be much larger. In the Apex Fusion work, he says, the team believes it can get to around 40 times the throughput of Cardano mainnet, with much faster finality.
That leads to the idea Hammond says they now want to deliver back to the community: partner chain in a box. The point is not only the protocol itself. It is the full toolset around it. Hammond says the goal is to let a builder press a button and bring up a basic partner chain in hours or minutes, rather than spending six months or a year doing the customisation from scratch. That, in turn, would create opportunities for wallet providers, explorers, tool builders and developers already trained in the Cardano ecosystem.
“You come along, you’ve got a need for a partner chain, and what you can do is just press a button, and within hours or minutes, you have a basic partner chain up and running.”
Hammond’s account also addresses a problem he describes with a memorable phrase: vampire chains. He points to Ethereum, where some partner chains pulled value out of the ecosystem rather than returning value to it. That is the risk he wants to avoid. His answer is complementarity. A partner chain should share incentives with the main chain, share participants where possible, and be able to reuse governance so that value continues to flow back into the ecosystem rather than away from it.
“The partner chain doesn’t suck value out of Cardano mainnet, but rather it returns value in a constructive way.”
That is the business case inside the engineering case. Hammond is not arguing for partner chains as a side route away from Cardano. He is arguing for a way to extend the code base, increase its utility and create new room for builders without weakening the main network. The protocol still has to be implemented, integrated, tested and tooled properly, and he is clear that this is the hard part. But the direction is set. The objective is to make Tachys the go-to mechanism for deploying partner chains that use the Cardano code base and to turn that into something that is usable by the wider community.
Hammond’s career gives that ambition some context. He moved from research into production systems, from language design into blockchain delivery, and from building new things to keeping complex systems alive over time. The common factor is engineering. In his account, the real test is not whether a system can be launched, but whether it can absorb shocks, maintain itself and keep producing value. Partner chain in a box is the next expression of that idea.
