Many actors are interested in running their business through blockchains as they are safe and resilient to failures.

Indeed, blockchains naturally mitigate insider attacks (malevolent system administrator, rogue employee, compromised employee account) since a single person can’t commit a transaction. Moreover, decentralization makes the system resistant to both technical failures (hardware issues, software upgrade issues, etc) and human-related errors (human errors or violence-related issues). Decentralization mitigates unforeseen threats such as outages or natural disasters. Lastly, there are many advantages to building on a blockchain besides those intrinsic to a blockchain’s design - we call these “incidental advantages”, and they include a strong and growing Web3 ecosystem, [list more advantages here].

However, typical blockchains lack the ability to tailor to specific needs such as:

  • Making a buffer available to companies with high NFT through put without committing every single operation on a major blockchain
  • Creating deterministic operating  systems
  • Enhancing an organization’s bus factor through decentralization
  • Freeing an organization from systemic risks (malvolent system administrator, compromised accounts, maintenance issues, etc) via a blockchain’s incidental advantages.

Unfortunately, creating a brand new blockchain from scratch yields a massive design, implementation and infrastructure overhead that most organizations aren’t willing to deal with.

All those specific needs can be addressed without a huge overhead by a single type of Layer 2: a sidechain, to which data and/or computations are offloaded.


A sidechain is a secondary blockchain that is paired to a parent blockchain and is able to communicate with it. A sidechain works like any blockchain and has the same core properties, with the added feature that its consensus is witnessed by the main chain.

To use a real-life analogy, think of a rehearsal studio who lends instruments to musicians as a parent blockchain. The band member who books a rehearsal slot and the required instruments is the sole responsible for all the instruments for the whole band. This band member makes sure nothing bad happens when the instruments change hands, takes care of them and is the one responsible for eventually returning them intact to the studio. In this situation, the studio doesn’t have the necessary means to oversee every details and defers the responsibility to this band member. This person can be thought of as a the studio’s sidechain.

Marigold action

Here, at Marigold, we address this through Deku. Deku is both an implementation of this kind of Layer 2 and the team that works on it.

Right now, the Deku team focuses on building a Tezos sidechain and the team’s long term goal is to create a framework for creating sidechains, allowing organizations to easily create blockchains tailored to their needs.

The main challenges are to understand, study, and implement various possibilities of known blockchain building blocks such as networking (typical problem: efficient communication between all nodes), consensus (defining a single consensus that works both in a private Proof-of-Authority setting and a public Proof-of-Stake setting), state transition (how to extract parallelism from a blockchain state machine), cryptoeconomics (finding the proper minimal values required to make the whole system run smoothly) while maintaining redundancy and a framework vibe in the architecture and code.

Team Deku is a diverse team of software engineers coming from a wide array of backgrounds, countries, and time zones. Marigold formed the team in July of 2021, and is growing the team rapidly. If you are interested in OCaml/ReasonML, programming languages, blockchains, and/or consensus, be sure to drop us a line!

Scroll to top