Inside the Machine: Doma's Tech Team on DAVs, On-Chain DNS, Limit Orders and the Road to DomainFi

In roughly seven to eight months since Mainnet, Doma's engineering team has shipped its own blockchain, an NFT marketplace, a launchpad, on-chain DNS management, Solana expansion, and a fractionalization protocol for real-world domain assets. And they say they are not even started yet. That is not marketing copy. That is the engineering record laid out by three members of Doma's technology org during the first ever tech-focused Discord AMA, which I had the privilege of hosting. As Kai Sung, VP of Engineering, put it: "It's hard to imagine that we only have been on Mainnet for like just seven to eight months."

I have been in web3 since 2016, and I have sat through more AMAs than I can count. Most of them are roadmap theater. This one was different. The conversation went deep into protocol design, DNS architecture, smart contract tradeoffs, and why bringing real-world assets on chain is fundamentally harder than building another DeFi fork. What follows is not a transcript. It is the architecture, the roadmap, the engineering decisions, and the vision pulled into a single narrative.

Three people from Doma's technology org joined me on stage.

Inder Singh, VP of Product and Tech, has been with Doma from the start. He comes from the domain industry and has seen both the pain and the opportunity firsthand. He is the person who can talk about why rides.com is on the platform and why only about 50 domain portfolios in the world are worth tokenizing, because he knows the people who own them.

Kai Sung, VP of Engineering, has been with Doma from the start alongside Inder. Based in the US, Kai leads the engineering org that architected the protocol bridging Web2 and Web3. He is the person who can explain why DNS has never gone down and why the team chose to retain admin control in certain smart contract functions.

Oswaldo Rivas, Senior Product Manager, joined the team in March and came from the Cardano ecosystem. This was his first AMA appearance, and he described the experience of working at Doma as "a roller coaster ride in the good sense" on a blank canvas of innovation. He is the person translating two fundamentally different user mental models, domainers and crypto-native traders, into a single product.

Before we got into the architecture, I wanted to establish something concrete: what has this team actually built and shipped? The list is substantial. They architected their own protocol to bridge Web2 and Web3. They built and launched their own blockchain. They shipped an NFT marketplace, a launchpad, subdomains, a quests feature with gamification, and on-chain DNS tooling that includes smart-contract-driven DNS management. They expanded to Solana with domains already minted on that chain, Solana wallet connection, and the ability to tokenize on EVM and bridge to Solana. They are applying for the .sol TLD on behalf of the Solana Foundation. They built both simple mode and advanced mode UI. And they shipped fractionalization of single domains.

That is a lot of surface area for a team operating in what Inder called the crypto bear market. "We're not even started yet," Inder said. "We're gonna go way, way, way bigger and way higher than this."

What struck me is that this is not a team chasing trends. Every item on that list connects to the same thesis: domains are real-world assets, they deserve liquidity, and blockchain is the infrastructure that delivers it.

The DAV, or Domain Asset Vehicle, is the next evolution of Doma's fractionalization model. Today, you can fractionalize a single domain into tokens and trade partial ownership. A DAV takes that concept further. It lets someone assemble a basket of domains, fractionalize the entire basket, and allow people to buy fractional ownership of the basket itself rather than individual domains.

Kai walked through the engineering. The team is building the smart contract support for managing a DAV and is targeting Solana for the DAV token, while using Doma chain as the source of truth. That means there is cross-chain communication happening between Solana and Doma chain under the hood. On the development side, Kai confirmed that most of the backend is already built, the indexing support is built out, and the DAV has passed its audit with Halborn with flying colors. Launch timing is being coordinated with the business side of the company.

Oswaldo framed the DAV as the natural evolution of what Doma has been building and described the creative possibilities once the initial concept is validated. He floated the idea of themed baskets down the line: a basket of AI-only domains, sports-related domains, web3 industry domains, or gaming industry domains. But he was careful to note that the team wants to test the core DAV concept first and validate demand, particularly with the Solana community, before pushing into themed variations.

The origin story matters here. Both Inder and Oswaldo emphasized that the DAV idea itself came from community feedback. As Inder put it: "A lot of what we're doing with DAVs came from people saying, hey, I don't want to think about ten different tokens. Is there an easy way I can get access to a whole bunch of domains?"

In Oswaldo's framing, the DAV experience will lean first toward the trader but will not leave aside the domainer experience.

When I asked Kai what part of Doma he is most proud of that users probably do not notice, he did not hesitate: on-chain DNS management. Doma has built smart contracts that let users change their DNS records and name servers directly on chain. That alone is significant. But the more powerful twist comes through fractionalization.

If a domain has been fractionalized and someone acquires more than 50 percent of the token supply, they gain DNS control over that domain. As Kai explained, not every fractional domain sells more than 50 percent of its supply, but for those that do, the majority holder can go and change the DNS.

The use cases this opens up are genuinely novel. A DAO could collectively buy a majority stake in a fractionalized domain and then manage that domain's DNS through smart contract governance. Kai also raised the possibility of domain leasing built on fractional ownership, where DNS control is temporarily delegated through token mechanics.

He acknowledged this is a feature the team should probably promote more often: "I would love to see more applications taking advantage of on-chain DNS with smart contracts."

This represents a genuinely novel intersection of Web2 infrastructure and Web3 ownership primitives. And it is live on the platform today.

The decision to stay fully compliant with DNS and ICANN rather than building an alternative naming system was made early and deliberately. Kai's position was unambiguous: "We wanted to make sure we're aligned with existing domain naming system ICANN rather than be yet another Web3 experiment creating an alt naming service. We don't think those alt naming services will survive."

His reasoning is grounded in infrastructure reality. "DNS has never gone down, ever. Ever since DNS was invented, it's never gone down. It's fully decentralized." He drew an explicit parallel to Ethereum's uptime record. There should only be one naming service to avoid confusion, and DNS is that service.

The hardest engineering tradeoff was bridging a Web2 centralized model with the Web3 ethos of ownership. When Doma distributes tokenized domains as NFTs, users naturally assume they have full, unconditional ownership. But because these are real-world assets governed by a centralized system, the smart contracts retain some admin control. For example, the ability to forcefully transfer a domain in the case of real-world disputes.

Kai was candid that this was a compromise on the decentralized model, but framed it as protection: these are real-world assets, and they need to be governed in the same manner they are governed in Web2. The team debated internally how far to push the Web3 ethos and ultimately decided to leverage blockchain for its clear benefits, instant settlement, transparent transfers, on-chain records, while keeping the domain lifecycle in sync with DNS backend systems.

I reinforced this framing during the AMA by comparing it to building on Ethereum and EVM rather than inventing a unique blockchain. Doma is not in the business of fixing ICANN. It is in the business of getting liquidity to domains.

I asked the panel where a well-resourced competitor would realize Doma is more complex than anticipated. Kai pointed to the collision of two deeply complex ecosystems that the team had to master simultaneously.

On one side sits blockchain technology. On the other sits DNS and a very antiquated protocol called EPP, the Extensible Provisioning Protocol that underpins domain registration at the registry level. Understanding how EPP works, how domain registration flows, and what the whole domain lifecycle looks like was a significant challenge.

It helped enormously that Doma built its own registrar, Interstellar, which Kai described as a guinea pig to test the full stack end to end. Beyond the technical challenge, there is the operational one: getting external registrars onboarded to integrate the protocol so that users actually have a path to tokenize their domains.

On the Web3 side, the team has effectively built every major category of decentralized application, a DEX, an NFT marketplace, a launchpad, but with domains as the underlying real-world asset rather than synthetic tokens. Kai noted that this combination of real-world asset complexity and full-stack Web3 infrastructure is both the hardest part and the most exciting, and it is a key reason engineers and candidates are drawn to Doma: "They are not building yet another crypto platform, they are bringing real money and real assets on chain."

Inder laid out a moat argument that goes beyond technology, and it was one of the sharpest takes of the entire AMA.

He acknowledged that with AI, the cost of engineering is trending toward zero. "Any company that says, hey, our moat is we have really good engineers, is gonna lose that moat pretty fast. But you will not lose your experience moat."

What AI cannot replicate is experience. Every person on the Doma team has either extremely deep experience in the Web2 domain industry or the Web3 industry, or both. Inder compared the situation to trying to run a stock exchange while knowing nothing about finance. The ability to write a high-performance trading engine does not mean you understand how domains work or how money-making in the domain industry works.

Then there is access to supply. Inder asked why no other platform has access to premium domains like rides.com, alert.ai, and smoothie.com, and answered that Doma's industry relationships are the reason.

He went further on the DAV side: "There are literally only 50 portfolios in the world that are worth buying, that have more than half a million domains. The only people who can actually tokenize these portfolios and bring the financing of the domain industry on chain is us."

Best team, best experience, best access to the best businesses in the world. In Inder's view, that cannot be replicated unless someone spent roughly 20 years in the domain industry, regardless of whether AI exists.

Inder took the scaling question with characteristic honesty: "We always say no bottlenecks until they show up." Each new spurt of growth brings new challenges, which he framed as good problems to have.

The team is actively preparing for the DAV launch by dramatically scaling up capacity. On the chain side, Doma works with Conduit, who has been helping with chain infrastructure. On the broader infrastructure side, the team is putting in more servers across the globe and is spinning up its own bare metal server cluster.

Inder noted that these are significant investments during what he called the crypto bear market. "This is literally the crypto bear market and there is no place on earth, I will bet you this, in crypto where you can get a reliable, stable asset that has been proven on chain ever."

You do not hear companies investing heavily in infrastructure in a downturn very often. But the team's conviction is that domains are a durable asset and the industry has been around forever, so the investment is warranted. They have also been talking to partners on the Solana side who understand how important the DAV is, and working behind the scenes to make sure the infrastructure can handle the growth trajectory they anticipate.

I closed the AMA by asking each panelist the same question: what are you most excited about that is coming next?

Kai Sung: Automated Trading and the MCP Server. Kai is most excited about the automated trading capabilities being built. Limit orders are already in testing. Users will be able to set a buy order for when a token drops to a target price or a sell order for when it reaches a target price, and the system fills it automatically. Beyond limit orders, the team is building more generalized trading strategies that users can subscribe to by creating a wallet and depositing funds. On top of that, they are working on an MCP server for Doma that will allow people to hook up their LLM and write their own trading strategies using the Model Context Protocol. Kai framed this as providing layered APIs so people can choose what level they want to build at, from simple limit orders to fully custom AI-driven strategies.

Oswaldo Rivas: The DAV. Oswaldo named the DAV without hesitation. He sees it as a natural evolution from single-domain fractionalization to basket-level fractionalization and is eager to test the hypothesis with the Solana community first. He floated the creative potential of themed baskets later but was clear that Doma wants to validate the core concept and confirm demand before expanding. He described the DAV as exciting both because the team has conviction in it and because the community feedback supports it.

Inder Singh: DomainFi. Inder is most excited about new ways to earn money on Doma, what he calls DomainFi. Today you can sell whole domains or trade fractional tokens, but he says that is not really DomainFi yet. With the DAV, holders will receive ownership proceeds as the domains in the basket sell, creating something closer to a yield-style product. Once those tokens exist, they unlock the broader DomainFi thesis: taking loans against domain-backed tokens, putting them in vaults, looping them in DeFi. He framed a future where someone might never buy a domain or even a piece of a domain and still tap into DomainFi. He said this vision gets realized in roughly the next six months.

Both Inder and Oswaldo described a tight feedback loop between the community and product decisions. Inder credited the community directly for the DAV concept itself. He also noted that a lot of the upcoming UI work, around finding your domains, seeing your positions, checking earnings, creating limit orders and order books, is driven directly by user feedback. Community members have personally DM'd him on Discord and X with valuable, specific feedback that has shaped what gets built.

Oswaldo described the challenge of merging two fundamentally different mental models, traditional domainers and crypto-native traders, into a single product. He compared it to a ride-sharing app that has to serve both drivers and consumers. He shared the lesson of the simple mode and advanced mode: the team built both to help different user types, but some users did not even realize they were on simple mode and got confused about missing features.

And then he offered my favorite analogy of the entire AMA: "You build a fork and then you give it to the user. You say, this fork is for eating pasta. And then you realize that people are actually using it to comb their hair."

The point is that users will always find unexpected ways to use a product, and the community feedback loop is how you discover that and adapt.

Kai issued a concrete invitation to builders. The team maintains extensive documentation at docs.doma.xyz, and there is a specific LLM-ready dump at docs.doma.xyz/llms-full.txt that you can feed directly into an LLM to brainstorm what you want to build.

He pointed to a previous AMA where Aaron, a developer, built a pixel grid demo, showing that anyone can register a domain, create a token from it, and use that token to build something interesting on the domain's website. As Kai framed it: "We're not just a cryptocurrency for domains, but a real world asset with a website that's browsable that you can host exciting and engaging applications on."

The Forge program is still open for external builders, though Inder noted the quality bar has risen because AI has dramatically changed how fast the internal team ships. If you come to Forge with a copy of something from another chain or nothing working, it will not clear the bar. But with AI tools, you can get much further and bring much more value.

Kai also noted that engineering team members are lurking in the Discord and are happy to help if builders run into issues. That matters. A good developer experience is not just documentation. It is knowing someone will answer when you get stuck.

A quick note on culture, because it tells you something about how this team operates. Doma is distributed across the US, Europe, South America, and Asia. They hold offsites roughly every six months to bring the whole team together in person. They met in Lisbon, Portugal in January and are now gathered in Poland, chosen for its central location relative to the team and its affordability. Inder noted that the biggest team presence is in Europe, making the EU a natural meeting point.

Oswaldo captured the vibe of building at Doma better than anyone: it is a blank canvas where you can create whatever you want, but also one where you regularly scratch your head trying to figure out how to solve a problem that nobody has solved before. Meeting the team in person in Poland, he said, was cool. Having the same debates they have online but face to face.

If you are holding Doma domains or fractional tokens, the DAV is the next major catalyst. It passed its Halborn audit, most of the backend is built, and the team is coordinating launch timing with the business side. When it goes live, it creates a new asset class on the platform: basket-level exposure to domain portfolios, with ownership proceeds flowing to holders as domains in the basket sell.

If you are trading, watch for limit orders and automated strategies. They are in testing now. The MCP server will let advanced users hook up an LLM and write custom trading logic. Kai described it as layered APIs: pick the level of sophistication you want.

If you are building, the invitation is open and the resources are real. Feed docs.doma.xyz/llms-full.txt into your favorite LLM and start exploring. The Forge program is live. Engineering team members are in the Discord. The thing that makes Doma different for builders is that every tokenized domain is also a browsable website you can host applications on. That is a canvas most web3 platforms simply do not offer.

And if you are watching from the outside, trying to figure out whether this is real, consider the shipping record. Consider that the team is investing in bare metal infrastructure during a bear market. Consider that there are only about 50 domain portfolios in the world worth tokenizing, and this team has the relationships to access them.

As Inder said at the close: "We're not even started yet."

Explore the platform at doma.xyz. Dive into the docs at docs.doma.xyz. And if you were not in the Discord for this one, make sure you are there for the next.