tl;dr
- Enterprises want to bring assets on-chain, but they need privacy.
- The current solutions, deploy your own private L1 or L2, creates fragmented liquidity, complex interoperability, and trust-based privacy.
- There’s a better architecture involving the Anoma PA, intents, and solvers.
Note to the reader. This is stream of consciousness. Claude 4.5 helped review and made some suggestions for feedback.
Products
One thing that the Anoma Protocol adaptor does well is it provides privacy to end users on existing EVM chains. In particular, the PA is deployed to Ethereum main-net where there is valuable shared state that can interact with the PA.
One challenge with building a general system that can verify resource machine transactions is understanding what exactly to build. Beyond that there is also the question of how to take what you build to market, as in the shape of a product.
A product is an item that is made or refined and marketed. For example, Aave is a product. Users interact with Aave either directly through smart contract calls or indirectly via third party frontends or mobile interfaces. AnomaPay presents itself as one such product. It’s an application that provides users shielded transfers built on top of the Anoma Protocol adaptor and is compatible with any instance of the PA deployed on heterogenous EVM chains. The application has a frontend user interface which may or may not be built and deployed by the application and infrastructure developers.
In the mind of end users which include businesses and consumers, a product then is often thought about in terms of what it does with all of its software components collectively. Without getting into linguistics, semantics, and grammatical syntax arguments just note when someone asks you what is x product you often will think of and respond with what it does. In the case of AnomaPay that would be shielded transfers of ERC-20 tokens. This product is most likely beneficial for consumers who want to use assets on existing EVM chains where they currently reside but are unable to achieve any semblance of privacy.
There are other types of products that often fall under the category of Business to Business or B2B. These are products for enterprises (new buzz word) or businesses to consume. Historically, Software as a service (SaaS) provides a good example of B2B products. Take Slack, a workplace communication platform for example. Slack specializes in selling subscriptions to organizations who need an internal messaging platform for communication and coordination. While Slack does have a free tier that individuals can use, its generally used as enterprise software.
Enterprise Blockchains bro
One recent phenomena in the crypto industry is the emergence of institutional capital. Institutions often traditional investment banks, hedge funds, private equity firms, and regional banks are all looking at bringing assets on chain. The tokenize everything narrative has caught on, and many of these folks are talking to various crypto startups about what services they can provide to assist with onboarding to crypto rails. You may be familiar with the “enterprise blockchain” expeditions and narrative from prior cycles. This narrative has re-emerged and institutions are looking to execute on this vision. Since that time popular blockchain architectures have transitioned from exclusively deploy your own chain to now deploy your purpose built L1 chain or build a permissioned Layer 2 chain. One nice thing about building a layer 2 chain if done properly is you don’t have to bootstrap a validator set, worry about paying token incentives for security, or decentralize your system.
As such one approach to enterprise blockchains is the Prividium product which is marketed and produced by ZKSync. A Prividium is a private Validium. A Validium is an L2 that keeps execution and data storage off chain but posts proofs on chain to Ethereum. The trust assumptions and security are not as good as a standard rollup where storage is on chain but execution off-chain, so the operator can freeze the L2 in the case of a Validium. However, in the case of enterprises they may want that level of control.
In this type of construction Privacy is achieved through trust and permissioning. The user submits a transaction to a private RPC operated by a Sequencer who the enterprise operates which controls the ordering and inclusion guarantees of transactions. Only a succinct proof of execution is posted to Ethereum to be verified. The proof validates that the state transitions were computed correctly, no invalid transactions were included, the operator followed EVM execution rules, and that the new state root legitimately follows from the old state root + transactions. However the proof does not protect against the operator doing anything shady like front-running, censoring transactions, withholding user data, or provide any ordering guarantees.
In this model each institution runs their own sequencer (transaction ordering), runs their own prover (ZK proof generation), controls their own data storage, decides when to batch and post proofs, and sets their own gas token (or none).
Privacy in this model is confined to data control and access locality, but not cryptography.
| Who | What They See |
|---|---|
| Prividium Operator | full transaction data, all balances, all parties |
| Authorized Users (via private RPC) | Whatever the operator’s rules permit |
| Other Institutions | Nothing inside your Prividium |
| Public/Ethereum | Only ZK proofs and state roots |
One challenge with this approach is that you will create many silos of capital which require interoperability to interact with any shared state on another chain. The entire purpose of tokenizing assets or building anything on chain is because you want to interact with shared state. Otherwise you might as well not bother with the additional costs of writing proofs to Ethereum. Also, this model requires a lot of trust assumptions for privacy and runs into various different liquidity issues. But most importantly you cannot have atomic compossability because privacy requires trust in this model. and so interoperability which is permissioned requires institutions to negotiate off chain about what assets to move and why.
To summarize briefly before moving on. In this “private L2” model
- Privacy = permissioned access + data not on public chains
- Trust = operator for liveness/access, ZK proofs for execution correctness
- Sequencing = each Prividium operator runs independently
- Verification = anyone can check proofs against Ethereum
- State = isolated, permissioned writes and interop
It’s essentially, “run your own private L2 with ZK succinct proofs settling to Ethereum” rather than a cryptographic privacy solution. The value proposition is institutional control and public opacity.
Can Anoma do Better?
Since the Anoma protocol adaptor is already deployed to Ethereum main-net, it can interact with valuable shared state that “lives” on Ethereum.
Ethereum provides strong security, censorship resistance, decentralization, and credible neutrality gaurantees. You can argue with these terms, but there really isn’t another chain outside of Solana or Bitcoin that is close here. Bitcoin is not easily programmable and Solana can be swapped for Ethereum in this example though there are differences with the above properties when comparing with Ethereum.
The Data Tells a Story
Ethereum also has the most value secured of any chain by far ~ 67.78% of all crypto TVS or $71.15B according to Defillama. In addition, It still retains the most developer mindshare, best tooling, most auditors, and largest ecosystem. These are some of the reasons institutions have chosen to issue assets on Ethereum rather than other chains.
Ethereum has $12B of non stable coin RWAs issued which is about 2/3 of all non-stable coin RWAs.
If you look at stable coins the contrast is sharp. Ethereum has 60% of stable coins market or $181.3 B of the $301.74B stable coins.
Whether you look at RWAs or stable coins, its abundantly clear that Ethereum has an edge. Since liquidity has a network effect, its likely this will continue for some time.
Where Anoma fits in this Picture
It is possible for Anoma to make a similar pitch to institutional clients as that of an enterprise chain (L1 or L2). Instead of building your own chain without cryptographic privacy but permissioned access and trust me bro privacy with fragmented liquidity, isolated state, and permissioned interop, these enterprises could just use the Anoma protocol adaptor.
In particular, enterprise customer could issue assets on Ethereum, move those assets via a custom forwarder contract into the Anoma PA. This enterprise customer could then stand up their own intent pool and solver. While the solver could attempt some private solving via complex cryptographic techniques which still need to be explored and further fleshed out, they could just run their solver in TDX. In particular, Intel TDX provides an isolated execution environment so the solver can process intents without exposing them, then output ZK proofs for settlement. The TEE provides solver-side privacy; the ZKPs provide settlement verification.
What’s nice about this approach is that each enterprise could build their own solver, a network of solvers, or use someone else’s. It’s entirely up to them. And since solving is done on demand they do not need to mint empty blocks per slot time as an L2 sequencer or L1 block proposer would.
The nicest thing about this approach is that an institution can interact with any state that is controlled by the Anoma Protocol Adaptor. This makes atomic composability trivial (note that in the future it should in principle be possible to compose Anoma PA state with EVM state synchronously). It also makes the system far more permissionless and private. In fact a user that is not interested in institutional trading or RWAs could still utilize an enterprises solver or intent pool if they wanted. All you would need is a bonding/staking mechanism for solvers who can be slashed for misbehavior a la the slow game.
In Prividiums, users get ZK proofs for execution correctness and pure trust for everything else (liveness, censorship resistance, ordering)
With staked solvers, users get crypto-economic guarantees. A retail customer interacting with an institutional solver doesn’t need to trust the institution’s good behavior, they have slashing recourse. This transforms “trust me bro, we’re a bank” into “we’re a bank with $x staked that you can slash.”
Beyond the single chain setting, institutions could now also have access to assets that are controlled by multiple protocol adaptors. The enterprise could take advantage of a future Anoma’s native controller system for asset movement and settlement.
Here is a chart breaking down some of the differences and advantages of the Anoma PA approach for enterprises vs. an enterprise blockchain approach (prividium in this example).
| Aspect | Prividiums | Anoma Protocol Adapter |
|---|---|---|
| Cross-institution tx | Chain A proof → Gateway → Chain B proof → Gateway → Ethereum | Single proof to Ethereum |
| Proving | One rollup proof per batch | One roll_up proof per transaction |
| Interop complexity | Merkle proofs, interop roots, broadcasters | Native, same state space |
| Liquidity | Fragmented across N chains | Unified shielded pool |
| Atomicity | Requires escrow contracts + interop protocol | Native, same execution |
| Infrastructure | N sequencers, N DBs, N RPCs, Gateway | Solver network + one contract |
| Latency | Multiple round trips | Single settlement |
Major
: the difference is that institutions get operational control via solvers processing their subset of global state, not via running entire separate chains.
Enterprises get better
- privacy
- interoperability
- asset compossability
- security
- decentralization
- credible neutrality
- and censorship resistance from Ethereum
- with theoretically lower operational overhead
Each enterprise would need
- to run a solver node
- geth/loadstar node
- intent pool (or share one)
- Solver has viewing keys for their institutional resources
- Solver whitelists which intents it will process
- Coordination for cross-institution transactions (solver-to-solver communication)
I’m also confident, based on conversations with folks in the industry, that the cost of running a solver is significantly less than the cost of running a specialized L2 sequencer and chain.
Conclusion
Anoma’s PA architecture which features a global shielded pool, permissioned solvers, information flow control (programmable disclosure), achieves the same privacy guarantees as enterprise chains on the market today with:
- Lower operational overhead
- Atomic settlement instead of multi-hop interop
- Unified liquidity instead of fragmentation
- State-level privacy instead of infrastructure-level isolation
The Anoma model is the right architecture for private, programmable, multi-party settlement for enterprises.