Disclaimer

This article does not contain financial or legal advice; it is just my personal experience as part of my desire to help and educate others. It is entirely separate from my current work at the Worldcoin Foundation or my previous work at Tools For Humanity.

TL;DR

• Angel investing is a great way to gain exposure to an industry in which you possess asymmetric knowledge or domain expertise.

• Advising projects for equity or becoming an investor because you are knowledgeable about the startup's technology, product, or industry is what I've found to be most lucrative and enjoyable.

• You are investing in people; remember that.

• Build relationships, build trust, and do things in earnest; your reputation is everything: how you make people feel and your actions to help them go a long way.

• Capital is abundant. Morality, integrity, intelligence, honesty, experience, wisdom, social capital, capacity to execute, and domain expertise are scarce.

• Look for projects that build positive sum products or services.

• Do everything the right way, don't cut corners, have a proper structure (accountant, lawyer, ...), and never burn bridges.

• I’m extremely illiquid for the next few years, this is not for people who cannot afford to lose the money they put in, everything is a bet

Introduction

I have been looking to write this article for a long while but never really got around to doing it justice. I'll probably have to write a follow-up to this article at some point, but I believe this is a good starting point. Please leave me any feedback on X or Telegram.

In late 2021, I wrote The Ultimate Guide to L2s on Ethereum, in which I explained how rollups work, my predictions for what Ethereum L2 and L1 scalability would look like, and the projects that were building said future. After writing it, publishing it, and posting about it on Twitter, I received a DM from L2Beat asking if I wanted to angel invest in their project since I was one of the most active community members in the L2 discussion space. At that time, I was just a 20-year-old who had made some money from airdrops and farming different protocols throughout DeFi summer. I had some money saved but didn't know what to do with it. I was willing to give it a try. What is the worst thing that could go wrong? It was a big turning point for me, now that I am reflecting on how that directionally impacted my life. It gave me an outside confirmation that something like angel investing was more accessible than I thought, as someone very young and inexperienced in investing. And for that, I will be forever grateful.

Since then, I have angeled in about 35 projects, and I still have some deal flow to go through that I will reveal once everything is closed and publicly announced. It has genuinely helped me a lot when it comes to having access to brilliant people working on the tip of the spear of each of their respective fields, and that is by far the most valuable thing that I have gotten out of angel investing. I have learned a lot in my journey, and I figured out that sharing with others could be helpful for those navigating this space because I sure know I would have appreciated something like this article existing back when I was first trying to angel invest.

One of the original motivations for writing this article is that, as a builder, you tend to overlook investing because you are already too busy with everything you need to do. However, engineers and builders have an edge over the market in terms of understanding many things. This is especially true in crypto, where many advantages are nuanced, subtle, and technical. A technical understanding of the space will give you a lot of edge in investing as an angel, and in general, it feels good to invest in technologies you believe in before they are available to the public market at a much higher valuation. Builders usually don't have the time or ability to manage their assets. So they get some ETH, maybe stake it, put some stables into vanilla DeFi for yield, and buy some liquid tokens of the projects they like. Still, there are only asymmetric risk/reward opportunities if you decide to spend a significant amount of time researching and investing. This is why angel investing is much more suited for engineers; it plays better into their strengths.

Early steps

It all starts with some form of luck, but luck is the byproduct of a lot of work. I spend lots of time on X/Twitter, and I have some visibility thanks to some of the articles and posts I have shared in the past. I mostly reshare what I like and occasionally publish a long-form article about something I care about like I am doing now. Visibility in the space around specific topics will make you part of the collective mindshare of the space. If the opportunity presents itself, people might contact you about various things, including angel investing in their companies or sharing deal flow between friends/acquaintances/friendly funds if they believe you are a good fit for whatever they seek. When I started angel investing, I didn't do any reachouts. It happened organically because the friends I met along the way at conferences happened to fundraise after I met them somewhere down the line. This is how I got access to my first few deals, and the only reason why I felt compelled to do so is because I wanted to support my friends who were building things they were passionate about and knew they were good at what they did. Wanting to support my friends in their dreams felt natural and aligned with my vision of who I want to become, so I started thinking about it more proactively after the first few occasions where it mostly happened in an organic way where my founder friends reached out to me.

Building a personal relationship with teams at conferences, deeply caring about what they are trying to develop, and being helpful wherever possible are the best ways to do it. Going to events and building relationships with many teams and individuals greatly increased my social capital over a relatively short time. One key thing to notice is that social capital compounds just as much as your financial capital, if not more, and I believe that social capital is the biggest driver in professional success in any endeavor, whether sales, building technologies, doing research or philanthropy/volunteering. It truly doesn't matter, having skilled friends in your networks who have other friends, they have capital, new insights, and wisdom, they are able to change things, and if you become friends, you can truly change the world together for the better. It's something that is better understood if you live through it. Having amazing friends and acquaintances is the proudest accomplishment in my professional life and a great source of joy.

The first few deals are a big learning experience, you have to figure out how you're going to do taxes, accounting, what is a signature block, how company valuations work, what pre-seed/seed/series A/B/... rounds mean, you start thinking about a framework for how you decide which companies to invest in and for what reason, you start forming your own opinions on the private market and different founders, you build relationships, history, you start advising some of them and try to help them succeed, some experiences may not be very pleasant and you start drawing lines for what kind of people you'd like to work with and on what terms. In general, there are so many things you can learn just by undergoing the process of angel investing in companies. My biggest takeaways are to get legal advice from an expert early on, set up taxes right (if you are a US person, F in the chat, but you usually have to be an accredited investor and other things), and just make sure you do things properly and you fully understand what you are doing. Signing a SAFE/token side letter without knowing what it entails is very easy. Having a basic education around them is very useful. If you have questions, try and ask friends who have experience and if you are doing due diligence as a founder, sometimes the funds who are leading their rounds or have participated in their previous rounds would be willing to help you. Or at least they are a good proxy for legitimacy in most cases since many funds have robust DD processes.

The angel game

As a builder, I was coding all day, but thanks to that, I also started understanding the different tech stacks of all the projects in the space. Being technical is a blessing, if you learn cryptography, for example, you can start reasoning about ZK/FHE/MPC startups, digital identity, blockchains, and so many other things. If you specialize in a specific subset of knowledge, you will start learning more about the tech than the market and, most importantly, than other investors. You can reason about which teams are doing a good job and which teams have a good technical roadmap, this will allow you to make an investment decision of your own will without having to follow narratives or other investors that you like. Because of it, you can make independent decisions and be early (pre-seed/seed). If you complement that with an understanding of how startups work, how products work, and the general lifecycle of the crypto market and different projects, you can start building an edge for yourself.

As a builder, I was tired of not having ownership over the things we were building together. Investors who provide capital arguably don't have as much insight as the people who are actually building in the trenches from the technical POV (of course, product, strategy, go to market, legal, … are where most investors help out), but this leaves a great and mostly untapped area of potential. Being able to help teams from a technical POV is my gateway to angel investing, but for you, it might be different. In 2021/2022, I understood where Ethereum scalability and AI x crypto were going, so most of the angel deals I did back then were in these areas. As a builder, you can provide a lot of value to startups by giving a small check-in and building a closer relationship, which can also develop into an advisor's role in the case of projects that you like a lot. A good way to understand how to get angel investing in dealflow is to understand the game and the reasoning behind founders wanting to have angel investors in the first place.

The pretext is that "capital is abundant," and therefore, it won't be the primary reason why you might be a good opportunity for the founder unless you can facilitate fundraising in the first place by introducing them to VC funds you have good relationships with and that they also like. There are different reasons: technical advisorships, connections to smart and useful people in the space, help with fundraising, help with hiring, narrative creation, go-to-market, social capital and reputation games, vibes, etc. In general, if you have experience in anything useful to help a startup succeed or provide value, you will be an interesting potential angel investor. I am not an expert in angel investing by any means, but I have some experience that I can speak to.

Most of my first angel investments were mostly because I met (crypto) friends at conferences who ended up becoming founders or were already at the time I met them, we happened to connect, and they asked me to angel invest in them, or I was excited about what they were doing so I asked if they were raising. After you get 5 or more angel investments done into relevant projects in your space, it also serves as a good track record and measures that you can be early into other companies that other people consider "hot." Therefore, if you have several "hot deals" under your belt, other "hot" companies will see that. It will definitely make the process easier over time, and you will even have reachouts from good companies as opposed to having to sweat and try hard to get them yourself. In the end, you have to do three things (according to Sam Altman, someone who led YC, his blog is a great resource if you want to become a great investor):

• get access to good investment opportunities

• make good decisions about what to invest in

• get the companies you want to invest in to choose you as an investor

Later stages

Now, after doing 35 angel deals or so and having about 11 more to fulfill in the next month or two, I think about how to provide value to my portfolio companies to help them succeed. One of the most exciting ways I like to work with my portfolio companies is to help them be synergistic, introduce the founders to each other, and help build synergies at scale. If there are many companies in the same vertical, then they can be each others' customers or have integrations, if there is a shared research/engineering problem, they can work on it together, they can build joint narratives and network effects, and have positive sum collaborations. Over the 6 years that I have been in crypto, I have managed to meet lots of people in the space and build a network of people that I'm quite proud of, and so what I like doing is introducing them to each other if they find any meaningful overlap. All of these compound effects of having a network are marvelous to watch unfold, synergies emerge out of thin air, and pieces I never saw coming start fitting in together. I can't wait to see what we're able to accomplish and build over the coming decades.

One thing that has helped me substantially is to network a lot with fellow angels and venture capital firms. Helping portcos from your favorite investors in various different ways and sharing dealflow with one another is a very positive sum. Let's say one of my friends is raising for a round. I can introduce him/her to 15-20 VC firms, and since I have a relationship with that firm, I can help with DD and act as a matchmaker of capital to startups. Some people make these arrangements in a more formal venture partner/associate format or on a per-deal-carry basis. I have explored such avenues personally, but I personally prefer to do it with no personal stake in it mostly because I don't want to manage the overhead of said deals, and it takes away precious time from other things. I'm happy to introduce my friends to my friends because I want to see both succeed, even when they happen to be VCs and startup founders. If you end up getting to a similar place and you'd like to spend more time managing your angel activities, then definitely explore such arrangements, as they can be quite lucrative and open a lot of doors. A substantial amount of my dealflow, especially recent deals, has come to me from my network of VC and angel friends through referrals as opposed to me having to reach out. Everything happened for me, intros, referrals, ... What goes around comes around.

My Portfolio

Clique (Seed, Series A)

I met Kevin from the Clique team close to the beginning of their journey at ETHDenver 2022. We met through Jessy's hackerhouse and spent lots of time together talking about a lot of things. At the time, I was looking for a job and exploring all the different possibilities, and Keving explained to me what TEEs, TLSN, DECO, identity attestations, and other things were. It blew my mind and heavily influenced my thinking, which helped me subsequently choose to start working on Worldcoin at TFH in July of 2022. I also met Jaden at Devconnect Amsterdam in 2022, and we have become close friends over time. Clique is building a general compute network with TEE-based TLS attestations (town crier extensions) for digital identity and various other use cases and has helped many different teams with getting insight on their users (Optimism, OffChain Labs, Eigenlayer, and many others). Definitely check out their docs to see what is possible with Clique.

Pimlico (Pre-Seed)

I met Kristof in Bogota during Devcon 6 in 2022 at a 0xPARC event where I brought an orb, he was super curious about it, and I ended up onboarding him to World ID. I explained how the World App worked using Safe, smart accounts, and meta-transactions (ERC2771), and he told me all about how he co-authored the account abstraction ERC and what Pimlico was all about. Pimlico has been building account abstraction infrastructure and has the most used 4337 paymasters and relayers, as well as an awesome OSS 4337 SDK called permissionless.js (docs). Kristof and his team are great people, and I always have tons of fun with him, either on calls, when I am around conferences, or in London.

Pragma (Formerly Astraly) (pre-seed)

I met Matteo and 0xevolve at a Starknet event in Lisbon in October of 2022. At the time, they were building a project called Astraly, which was building on-chain reputation primitives and a reputation-based token distribution platform on Starknet after the original Empiric Network team wound down, the Astraly team took over and created the Pragma Oracle network, which utilizes STARK proofs to ensure data correctness while maintaining low-latency and composability. They were one of the first teams to convince me that there was a great developer ecosystem on Starknet and taught me about Cairo and its evolutions. They are core contributors to Madara, which is a rollup stack using the Cairo VM using the Substrate node stack (now just called Polkadot SDK), which will be used to power the Pragma V2 oracle network (watch out for announcements on the Pragma X).

Giza (Pre-Seed)

In October of 2022, I started working on proto-neural-zkp, a prototype POC to benchmark the current state of the art of ZK at the time (plonky2) in order to prove a simple neural network. I implemented this demo and learned Rust in the process, and around this time, I started learning more about other teams who were thinking about this problem space, so I created the ZKML community Telegram and the awesome-zkml community resource aggregator and filled it with everything I had found until then. In the Telegram group chat, I added everyone I knew at the time who was working in the area, and they themselves added other people who knew, and that's how the community grew to its first 100 participants. During this time, we organized a couple of online meetings, and that's how I met the teams behind Modulus, EZKL and Giza in the first place. After that, I met Fran and Renç in person in Lisbon for ZK Summit 9 (and the ZKML side event) and in Lustica Bay for the first iteration in Zuzalu, and I ended up investing shortly after that. Giza is building actionable AI solutions for dapps, which allows you to define your own AI actions to be executed on/off-chain, as well as the Orion Cairo library, which allows you to create validity proofs of the most popular AI/ML algorithms.

Modulus (Advisor equity - Pre-Seed)

I met Daniel, Ryan and Nick first in the ZKML community telegram group chat and over the online meetups we had. The first time we met in person was at SBC in Stanford in 2022, and later at the first-ever ZKML community meetup during ETHDenver 2023. We bonded a lot during the event, and I decided to become an advisor for the project in exchange for a small amount of equity at the seed valuation. The friendship I've built with the Modulus team developed into a closer partnership with Worldcoin, where we are leveraging their ZKML and applied cryptography expertise in order to build World ID in a way that Worldcoin will never touch the plaintext of an iris code ever again (still WIP). Leveraging Modulus' state-of-the-art GKR + LogUp proving system called Remainder, we are exploring the creation of a client-side ZKML prover on edge devices (phones that have the World App installed) in order to make the creation of the iris code purely self-custodial and re-registration to World ID permissionless (if you ever update the iris code pipeline, the output space changes thus triggering all users to have to visit an orb again, this approach would prevent that). The Modulus team has focused on fundamental cryptography and took a first principles approach to building ZKML infrastructure, which is one of the reasons I'm super excited to work together with them at Worldcoin and be an advisor to the startup (plus, the team is awesome).

Ritual (Series A)

I met Akilesh through a Telegram introduction/referral from the Modulus team when Ritual was raising their Series A and was looking for good angels. We immediately clicked and shared a very similar vision for what needs to happen in order for AI to become more decentralized and commoditized for web3 builders. Ritual is building a lot of tooling that allows you to consume AI models on-chain in a verifiable way while being able to choose your own cost/security tradeoffs for the proofs (cryptoeconomic - restaking, TEEs, OP, ZK, ... by specifying a Verifier in the request) using InferNet. To learn more about Ritual, I recommend listening to the ZK Podcast episode featuring Ritual. They are still early on in their journey, but I'm super excited to see what developers will build with their tooling, and I can't wait to get my hands on some of their tools once I have some time for it.

Nebra (Pre-Seed)

I met Shumo during Devcon 6 in Bogota near one of the ZK events in the main conference buildings, at the time, he was still working at Manta and exploring new research directions. Over time, he settled on proof aggregation and built Nebra's Universal Proof Aggregator (UPA) with his team. For now, UPA only supports groth16 proof aggregation, with more proof types on the way. This is something that would be really useful to Worldcoin since every two weeks, we submit about 3M+ World ID Semaphore groth16 proofs, which are claims of the biweekly Worldcoin token grant for all World ID-verified users. It creates a big blockspace footprint on Optimism mainnet, and therefore, proof aggregation would be great to batch these proofs off-chain and put them on-chain once the aggregate proof is verified in the aggregator contract, which would heavily amortize and alleviate proof verification gas costs. This is why Nebra became a Worldcoin Foundation grantee for our first wave of grants. I'm excited for when heterogenous proof aggregation (proofs using different proof systems) comes along so that we can create bespoke composite proofs of different kinds (storage proof + web proof + co-processors + Semaphore + zkKYC + many others). The Nebra team is great, and I always enjoy spending time with them at conferences. Check out their Proof Singularity events if you can go to them (there will be one at ETHCC; this was the one in Denver).

Pluto (Seed)

I met Tracy for the first time in person at ETHDenver 2023 at the ZKML meetup. He had already been building Pluto at the time, but he was considering different ideas to pursue directionally with the startup. He ended up converging on web proofs, which allow you to make attestation on arbitrary web data using Pluto's platform, leveraging the power of the TLS protocol and ZK proofs and building many other forms of novel cryptographic tools for application developers. In order to learn more about what Pluto is all about, I recommend listening to the ZK podcast episode talking about web proofs with Anna Rose and Tracy.

Web proofs are an integral part of my future of digital identity thesis and it's going to enable much more than that. Excited to see what Tracy and his team will build in the future. If you want to try early access to their platform, you can sign up for this form.

Eclipse (Series A, strategic round)

I am very interested in the scalability roadmap for Ethereum, and I believe that the Solana ecosystem of applications has a lot of cool things to offer. Eclipse is building a Solana Virtual Machine (SVM) rollup on Ethereum. I was interested in learning more about the SVM at some point and what applications are being built on Solana, through my conversations with Jon from DBA (one of my all time fav lower-case-r researchers in the space), I found Eclipse to be a really cool middle ground, so I reached out to their team to angel in the Series A and their researchers/engineers round. I will always try my best to support cool experiments on Ethereum around scalability and onboarding different kinds of developers onto it. If you want to build on the SVM and deploy on Eclipse, check out their docs.

Friend (Formerly Tab) (Seed)

Jessy's first-ever hackerhouse during ETHDenver 2022 was a pivotal moment in my life, I met so many cool and talented people, as well as the Alongside team, which I ended up working with for a few months. One of the people I met there was Avi, who at the time was a student at Harvard and he had built the most popular covid 19 tracker website. After ETHDenver, the war in Ukraine started, and Avi built a non-profit matchmaking platform for refugees and people who were offering their homes called UkraineTakeShelter (no longer active) and another humanitarian effort website around the time that earthquakes hit southern Turkey. I was always impressed with Avi, and when he came to my friend Miguel Piedrafita and me to talk about how he was super excited about modern LLMs and turning them into wearable companions you can bring everywhere, I was super excited to learn more. He flew over to Lisbon and stayed at Miguel's for a few days, we hung out, and they both showed me the prototype (for which Miguel built all of the backend plumbing and AI API logic with Avi in SF), I was instantly sold on the idea and when Avi announced his seed raise I was super excited to support him as a friend. Back then, it was called Tab, but Avi bought the Friend.com domain and rebranded the project. I can't wait for the project to launch on July 30th and get ahold of my own Friend AI wearable necklace.

OpenQ (Seed)

Another cool person I met in person for the first time at ETHPrague 2022 was Rick. At that point, I had been working at Alongside and exploring a bunch of other areas in the space. Since I created devpill.me, which is an OSS public good blockchain development guide (out of date at the moment), we started chatting about developer education, developer relations, and the adoption of tech in the space. He had just started his startup OpenQ, which at the time was a bounty platform and hackathon infra service platform, which eventually pivoted to being a devrel CRM, which is what it is now. In the crypto space and beyond, the scarcest commodity is usually quality developers building interesting applications with your technologies. In crypto especially, many teams pay millions of dollars for conferences, educational programs, grants and many other opportunities in order to be able to attract developers to their ecosystems. However, these efforts are really hard to measure objectively, and it's hard to see the value per $ spent. OpenQ is positioned to disrupt the devrel space by providing a fully featured and data-driven solution for managing your devrel workflows.

Monad (strategic round)

I met Keone at ETHDenver 2023 at a Monad brunch. At the time, I didn't know that much about what the team was building, but over time and at many different events and chats, I discovered the plethora of interesting research directions and understood the tradeoffs that the Monad was taking in order to scale at the L1 level. I love Ethereum, but I also like teams that are trying to innovate at various levels of the stack, not only rollups. Many people dislike new layer 1s because of what they have entailed historically, some of them failed to garner ecosystem adoption, some of them have no major innovations and are just hard forks of existing toolchains, and some of them have been disingenuous about the tradeoffs they take, and some were pure cash grabs. What I like about Monad is that they are building their entire node infrastructure from scratch and have innovated on everything from first principles. Of course, the decentralization tradeoffs they have taken are worse than Ethereum L1 because the hardware the node runs on will have to be quite performant and will have to have a similar profile as a Solana node, but they are working on a lot of the current problems like parallelization of the EVM, better merkleization, better I/O and p2p pipelines and a bunch of other things that have become quite stagnant on Ethereum mainnet because it's focusing on improving on the rollup oriented roadmap to become the best data availability layer while maintaining the highest levels of decentralization and economic security. My opinion has changed over time on L1s, and I am much more appreciative of the exploratory work teams like Solana and Sui/Aptos have taken because it's still worthwhile to explore these routes of scaling. If you zoom out to a decades-long time horizon, anything without merit will die out anyway, and sustaining a dead ecosystem will be hard. However, all the lessons learned along the way will always be valuable.

Herodotus (Seed)

I met the Herodotus team at a Starknet event in October of 2022 in Lisbon, at the time, I was fairly new to what storage proofs were, and they explained a lot of these new concepts to me. I met Marcello and Tiago (who has moved on from the team recently to explore new directions), whom I became friends with because they were both living in Lisbon, just like I was at the time, I also met up Kacper and Pia while they were visiting Lisbon. At the time, they were raising their pre-seed, and it was the bear market. I didn't have any spare capital to allocate, but I ended up investing in their seed round last year. Herodotus, to me, unlocks a whole new world using storage proofs. Seamless cross-chain bridging of assets or general state snapshots that you can feed to co-processors like SP1 or Risc0 to perform computations in a completely provable way is an extremely powerful primitive and powers protocols like SnapshotX for being able to vote with your tokens from any network on any network (Ethereum L1 and L2s with no added trust assumptions).

OnlyDust (tiny friendly check)

I only put a tiny check into OnlyDust, but I like them so much that I couldn't skip them. I met Greg from Only Dust on X a long time ago and in person at Devconnect 2022 in Amsterdam. He's one of my favorite people in the space. OnlyDust is building a platform to match developer ecosystems to OSS contributors. They have worked with the Starknet, Cosmos, Optimism, Ethereum, and Aptos ecosystems. A bunch of really cool projects got a ton of great OSS devs working on them. If you are an open-source developer and you don't know how to get funded, then take a look at OnlyDust's platform, and if you have a developer ecosystem, consider putting up bounties and grants to fund OnlyDust contributors to work on your OSS projects.

Movement (Series A)

I have mentioned Ethereum scalability and exploration in this domain a couple of times, so this won't be a surprise. I got introduced to Rushi through a mutual friend over Telegram and met up with him recently during Token2049 week in Dubai at a HadronFC meetup organized by the one and only ChainYoda. Movement is building a MoveVM-based rollup on Ethereum and enabling the Aptos/Sui ecosystem of applications to live on Ethereum. I still have to learn more about the Aptos and Sui ecosystem and Move, but as I mentioned many times in this article already, I believe in experimentation. It's also why I like projects like Starknet with Cairo, Fuel with Sway, and others. It's the only way that we don't get stuck with local optima. If you want to build on Movement check out their docs.

Prime Intellect (Seed)

Prime Intellect is one of my favorite projects that have emerged that work at the intersection of crypto and AI. I've known Vincent for almost three years through his work in the DeSci space with MoleculeDAO and VitaDAO, he and a bunch of other people originally pilled me on longevity and biotech research. I've read a lot of the resources he's sent me over time, and I have learned a lot from him. When he told me he had moved on from DeSci in order to help decentralize AI hardware, I was super excited as I had been working on ZKML-related community work for a while and was wondering what other ways crypto tech can be useful in the AI domain. Marketplaces for hardware almost seem obvious in hindsight, but no team executed this vision properly, maybe apart from teams like Akash (io.net, for example, seems to have lots of controversy, and BitTensor seems to lack productive use cases). Prime Intellect democratizes AI by computing by providing an aggregated interface over all of the major clouds and third-party compute providers, ensuring users get the best GPUs for the lowest prices and helping decentralize AI over time, especially as more independent compute providers ramp up.

Mode (Series A)

James from Mode was introduced to me by Vincent, and I really liked that there was a team trying to experiment with sustainable incentive structures for the Superchain ecosystem because until that point, Arbitrum had an unquestionable dominance in DeFi apps and usage and there wasn't a clear strategy in the OP ecosystem to really challenged that outside of apps like Velodrome, and a few others. Mode is an OP-stack L2, which is part of the Superchain and is exploring new economic incentive mechanisms for DeFi. It is compelling to me in a world where most L2 teams are hyper-focusing on infrastructure.

Praxis (Series A)

In May of 2022, I went to the first iteration of Zuzalu, and I spent a whole month in Luštica Bay in Montenegro. While I was there, I met many cool people working on different problems and people interested in the emerging idea of network states pioneered by Balaji, Vitalik, and several others. Zuzalu was a great first experience of what such a network state might feel like. A cohesive group of people working at the tip of the spear of their fields having conversations and working together from the same place, creating value in a co-located environment, sharing ideas, and spending time together sounds like almost too good to be true, but that was exactly what it felt like. That is exactly what Praxis is trying to create. First, it is bootstrapping an internet-native group of founders, researchers, engineers, creators, and builders of all kinds and trying to create synergistic events in this network and going to different sovereign nations with interesting proposals to create special economic zones within their countries where Praxis members could create their business, communities, and network cities on top of. I met Dryden through a mutual friend introducing me, and we hopped on a call after 30 minutes. I was immediately convinced of their vision, and I immediately applied to become a Praxis citizen and got my Praxis Visa. Shortly after that, I participated in their Series A and advised them on blockchain scalability and digital identity from time to time.

GasHawk (Seed)

I got introduced to the GasHawk team (Dan and Daniel) by a mutual friend, and after getting on a call, I thought to myself, why haven't I ever heard of this, I could have saved myself a ton of money. GasHawk essentially helps individuals and teams alike optimally place their transactions on-chain in order to minimize their gas costs by submitting transactions through their RPC at periods of low blockspace demand. This infrastructure should be part of every single relayer and user pipeline that is not time-sensitive. I invested in their seed round and started using the GasHawk RPC for my personal transaction placements to mitigate my on-chain gas costs, which over the last 5 years have grown to be over 10 ETH in total gas spent.

Rhinestone (Seed)

Rhinestone also got introduced to me by a mutual friend that I share angel dealflow with. I met Kurt and Konrad over a Google Meets call, and I really liked their vision of an app store for Smart Accounts. At Worldcoin, I've had to work quite a bit with Safe Smart Accounts and their module ecosystem, and I have come to appreciate the quality of life improvements things like passkeys, account abstraction, and custom logic modules bring to the smart account space. We use a vesting module as well as a recurrent transaction spending limit module, and I have seen just how hard they are to build, curate, get the right APIs in, audit them, and then release them to the public. Many different teams build these modules for different purposes, but they lack discoverability, standardized APIs, auditing standards, and much more. This is where Rhinestone comes into play with its modular account abstraction infrastructure. To learn more, I recommend checking out their documentation.

Gevulot (Seed)

One of my VC friends introduced me to the Gevulot team, who participated in their seed round. I met Teemu over a meeting, and he gave me a lay of the land of proof marketplaces, where each project stands, and what their strategy is to capture market share, after which I was convinced to angel. They have been doing really well getting OSS contributors to get modules for different provers integrated into their live tesnet deployment. Several people at hackathons, like ZK Krakow, submitted projects as an integration of a new proof system into the Gevulot prover network.

Firstbatch (Seed)

I met the Firstbatch team by pure chance. One of my followers on X reached out to meet to me in person while I was in Istanbul visiting my girlfriend, and I agreed to get lunch with him. We immediately hit it off and became friends, and he wanted to introduce me to his friends Ömer and Efe, who are building FirstBatch. FirstBatch is the team behind Dria, which is a marketplace for synthetic data that AIs can use for finetuning and RAG. They are building many innovative solutions and doing fundamental research to solve the data problem in AI in a decentralized and open-source fashion where participants are fairly incentivized while doing so. They have lots of things coming in the next few months. Also, they are big One Piece fans, which immediately sold me the fact that they are wonderful people, and we became great friends over that really quickly.

GatlingX (Pre-Seed)

Eito was introduced to me by a mutual acquaintance, and I was really impressed when he showcased their GPU-EVM implementation numbers. The GPU-EVM can be used to parallelize transactions and to simulate the execution of the EVM for security tools like fuzzers for DeFi simulations (e.g., arbiter) and much more! You can read more about the GPU-EVM here.

Succinct (Series A)

I first met Uma on a call around using their Telepathy ZK-based bridges to bridge World ID to different roots, and I was very impressed with the stuff they were building. We hung out at ZK summits and chatted over DMs since and had lots of fun talking about ZK. Ever since they released SP1, and have been working on cool things like Celestia BlobstreamX, which made me very optimistic about the state of the ZKVM space. I angeled in their Series A because I really like their team and their vibes. There are a lot of competitors in the ZKVM space, and all of them are building great tech, but I like the OSS vibes that Succinct has, where several people have tried to build cool POCs with SP1 and the pace of development feels very fast.

Astria (Series A)

I was introduced to Josh by a close friend of mine who is a VC at a local Czech VC. I had been aware of Astria for over a year prior to meeting Josh on a meeting, and I had interacted with their team a bunch over time on X. Astria is building a decentralized, shared sequencer network to bridge the interoperability gap in the modular scalability paradigm. There are a lot of interesting research directions in MEV, interoperability, and decentralization of the modular stack, and I’m happy a talented team like Astria is working on this.

PWN (Seed)

PWN was founded by a really good friend of mine I met at my first crypto event eve, NFCastle Prague, which was held at the Prague castle in 2021. I subsequently helped Josef, the PWN team, and the DuctTape s.r.o. team with organizing the first ETHPrague event in 2022; and that is Josef. PWN is a p2p lending market that allows lenders and borrowers to define loans on their own terms, use any token or NFT as collateral, and specify custom repayment schedules with no price-based liquidations. The V2 of their protocol is on the way, and I can’t wait to start using it.

Berachain (Series B)

I was introduced to Smokey by a common friend. At first, I didn’t really understand the concept of Berachain beyond it being an EVM Cosmos SDK/Tendermint chain until I better understood the idea of proof of liquidity and how you can align incentives between validators in a way that allows you to align protocols and validators on the network. If the idea works in practice, it’d be a very innovative crypto-economic tool, and I’m all for it. Not to mention that the Bera ecosystem vibes are only matched by a few.

Fabric Cryptography (Series A)

I met Michael this year at ETHDenver, and it was one of my biggest learning experiences. Michael is an expert in hardware acceleration and has taught me a ton about the hardware industry and how the entire process of developing your own chip works. Fabric is building a verifiable processing unit (VPU), which is a specialized processor to perform cryptographical operations and, most importantly, in the realm of ZK. I really like the set of tradeoffs that Fabric is taking and the performance they are able to get on their V1 prototypes. Hardware will speed up the ZK industry by a lot and make a lot of use cases possible on the server proving side and also on the client side proving side with a lot of really interesting delegated proving schemes using MPC and ZK that are coming out.

Phylax (Seed)

I met Odysseas at ETHDenver in 2022 when I was pretty early on to my journey. At the time, he was consulting and contracting with various teams, and I remember learning a lot from him about Rust, the crypto space, other research topics, and life and philosophy. Odysseas soon after joined Nomad and went through the awful experience of the protocol getting exploited, which gave him the idea for Phylax (Guarding in Greek), which is a network that can actively monitor and censor malicious transactions in the mempool if they violate a set of invariants. These guardrails and policies exist in most distributed systems but have not yet emerged in crypto because of censorship-resistance guarantees. However, if we agree on a set of rules or invariants that can’t be violated on an app-chain or through a form of decentralized governance, these guardians will help us to mitigate lots of potential exploits that could happen in the future. I’m really excited about all the tooling that Phylax is building, check them out on GitHub!

Aligned Layer (Seed)

Aligned Layer is a project that is incubated by the LambdaClass team which I’ve known for several years at this point. I have always been friends with them and looked up to their work. I met them in person at ZK 9 in Lisbon at a ZKML side event, and that is also where I met Fede and RJ. Aligned Layer is a network secured as an EigenLayer AVS, which verifies ZK proofs quickly and cheaply. The Aligned Layer whitepaper goes into much more depth as to how this is accomplished, and I also recommend testing out their testnet. Learn more about Aligned Layer in their documentation.

Morpho (Series B)

I started using Aave when it was still called EthLend, and I've seen the market maker and lending markets evolve a ton throughout the times, but over recent years, I haven't seen that much innovation apart from Spark and a few other cool lending markets. Morpho has reopened my eyes to what a lending market could become, and I was more than happy to support them with a small check during their Series B round.

Morpho is building a permissionless and un-biased non-custodial lending protocol, where lenders and borrowers can define their own terms and collateral types.

I'm looking forward to seeing what the DeFi landscape will evolve into in the next few months.

Atoma Network (Seed)

When I lived in Lisbon, I met with a lot of people from the local community, and Jorge Antonio was one of them. We happened to have a shared group of acquaintances, one of whom I used to advise on Solidity for their blockchain developer boot camp at the time. Jorge is a friend of one of the founders. Atoma Network is a decentralized, verifiable AI inference network that allows you to query arbitrary models and get proofs of inference on arbitrary data. It’s building in a similar direction to Ritual but with several different approaches, and I don’t believe they overlap. Even if they did, the TAM for decentralized, verifiable AI is much too big for one single project to tackle.

Nameverse (Seed)

My best friend introduced me to the Nameverse team (Dennis and Mike) after coding with them at a hackathon, and I was really impressed with their vision for the world of namespaces. Nameverse is building Smart Domains, a multi-chain namespace protocol on top of ENS, simplifying smart wallet onboarding. In my many conversations with the team, I can see that they have a product sense like very few other crypto companies, and they can execute their vision almost at the speed of thought. This impressed me again and again, and I’m super proud to advise them on their journey. ENS built the protocol and the interfaces and helped scale it to multiple chains. However, we still lack a coherent product that expands on all of that and drives the UX and feature set home, one that is easy to integrate for all app developers alike and one that allows builders and users to auction off namespaces according to their own rules and incentive structures.

Jokerace (Seed)

My CT friend David Phelps has been writing tweets and long-form articles about different problems for about the same time that I have, and we have discovered that we share very similar views when it comes to a lot of things. Jokerace is building the primitive to power and monetize onchain social interactions through contests to become the internet's decision layer. I’m super excited about this primitive being used more to determine outcomes and actions to take onchain.

Nillion (Series B)

Nillion is building a blind computer that allows you to run computations on top of your data using a variation of multi-party computation (MPC) called nil-message computation (NMC). I am really interested in the modern programmable cryptography landscape utilizing ZK, FHE, MPC, and TEE primitives, and the Nillion team is building scalable infrastructure to address the problem of private computing. The valuation was a little high since it's already a Series B, but I really like their vision, and the possibilities for private computing are endless, if they build a platform and developer tooling that people like, it will bring countless use cases closer to being real and usable products without sacrificing user experience. I met them fairly recently, but so far, I've really liked my conversations with them, excited to learn more about the project and see how it develops.

PIN AI (Seed) (in stealth)

This team is still in stealth, so I won't share who is behind it, but I can share a bit about what they are about. They will tap into the power of modern LLMs and local data in an open-source way, where the user owns their data. We have seen the WWDC presentation recently, which introduced Apple Intelligence, which is a closed ecosystem where even though most inference happens locally, we don't know what is actually being run, and we can't connect arbitrary data to it unless Apple releases an integration or delegates it to a third party API like ChatGPT from OpenAI. They will come out of stealth in the coming weeks, so stay tuned. I also have a similar project in the pipeline, which is targeting a different problem space within local inference and personal data from all of your different applications.

Narratives I'm excited about

From my portfolio, you can sort of tell what the narratives I'm excited about are, but I'd like to expand a bit as to why I believe in them. The first major category for me would be scalability, at Worldcoin, I have seen true scale, and I see the issues we face all the time due to lack of blockspace, slow block times, slow throughput, state bottlenecks, proving costs, and more. There are a bunch of interesting approaches to solving different parts of the equation. I missed the entire set of rounds for the major rollups today, but there are new takes on different parts of the scalability equation: Astria is trying to solve shared sequencing and fragmentation, Eclipse and Movement are trying to bring the Move and SVM ecosystem of developers over to Ethereum which can help their applications settle on Ethereum as well, Nebra and Aligned Layer are solving the onchain verification costs of ZK proofs on the EVM (Worldcoin is 45% of OP mainnet, most of which is groth16 ZK proofs verifications), Fabric Cryptography is building ZK ASICs to reduce prover cost (this will reduce ZK rollup proving costs and drive down costs for users), Monad is building a performant EVM L1 to try and push experimentation at the base layer to the maximum (a lot of synergistic trends with Ethereum for R&D and cross-pollination), Phylax is trying to help mitigate exploits at the rollup execution level, Succinct allows you to offload computation from the chain and more. There are 2 scalability-related projects that I have in my current pipeline, which I'll announce on X in due time once the rounds close and become public.

The second category is the intersection of AI and crypto(graphy). Before I was into crypto, I interned at an AI startup doing computer vision. I was building with OpenCV, Pytorch, and the usual set of scientific computing libraries in Python (pandas, matplotlib, numpy, scipy, scikit-learn, etc). I was really excited about the topic and have been keeping up with it to this day. The reason why I chose to specialize in crypto as opposed to AI is because the ideals and culture of crypto match my own much better. I am a proponent of individual sovereignty and hate the way most systems we rely on every day are architected. In the AI world, if you want to work at the tip of the spear, you need to move to the US, join one of the top MANGA companies or OpenAI, or become a postdoc and researcher at some university or company research department. You lack sovereignty or freedom to choose your own research domain unless it's part of the "hot" set of topics, and you can get funding for working on an idea, and the terms under which you work don't give you full freedom of expression. All of this secrecy and making pretend games in public really gave me a bad taste of the web2 and AI tech sectors, so crypto was much more interesting. Not only is the technology extremely useful and interesting, but the people in the industry embody the values I do (for the most part). However, in my mind, I still believe AI is the single most important and exciting technology possible in this century. What if we could disrupt the status quo of AI and leverage crypto rails and tech to help it become more cypherpunk? Well, that is my crypto x AI thesis in a nutshell. In 2022, I discovered ZKML through my work at TFH on a cool experimental idea around ZK, proving the creation of an iris code client-side using ZKML, and I ended up creating the ZKML community (more here) where I met most of the people working at the intersection of AI x crypto. Ritual and Atoma are building verifiable AI inference networks and oracles, Friend is building a user-friendly LLM-enabled wearable pendant, Giza and Modulus are building ZKML tools and infrastructure, Prime Intellect is commoditizing AI hardware and building marketplaces on crypto rails, FirstBatch is decentralizing the creating of synthetic AI datasets and LLM executors, Nillion is helping AI inference become more private (I also really love Bagel Network and Zama, but missed their rounds), PIN AI is trying to tap into local LLMs with client-side inference on personal data, and I also have one other unannounced AI startup in the pipeline. There are so many exciting things happening in this field, and it makes me so happy that I'm able to rekindle my love for AI without having to compromise on my values.

The area that I'm probably most excited about is digital identity. Throughout November and December of last year, I wrote my "The Future of Digital Identity" blog, and I talked about all the possibilities of the current digital identity landscape, if you don't want to read the blog, you can watch the talk I gave at ETHPrague recently. Clique started as a company that allowed web3 applications to get web2 user information through a combination of TEEs, MPC, and ZK and has expanded into a general computation network with coprocessors and integrations with Axiom and Succinct, Pluto is a web proofs platform - it allows you to fetch web data using zkTLS and then make attestations about it using a Solidity co-processor, Herodotus is all about storage proofs - fetching on-chain data with provable provenance and being able to make attestations about it is an indispensable identity primitive, Praxis is building an internet-native nation, and JokeRace is building contests to power and monetize social interaction onchain. There are so many interesting things happening in the digital identity space and I think it will be one of the most interesting source of interesting use cases in 2025 and will be exacerbated by the 2024 presidential elections in the US.

There are other narratives like account abstraction, tooling, infrastructure, and DeFi that are super interesting but are a bit further from my expertise. Here, I made bets on teams I believe in, and I believe that these spaces will enable a lot of innovation. But I can't paint a compelling narrative without important context. My portfolio companies in these areas are Pimlico, GasHawk, Rhinestone, OpenQ, PWN, Berachain, Mode, Morpho, GatlingX, and the majority of the remaining deals I have in the works.

Some friendly advice (NFA)

One of the most non-straightforward things about angel investing, especially if you do it often, is the communications and administrative overhead it creates. To manage communications, I use a system of tags and reminders on Telegram (using premium). I have different groups for different kinds of messages. If it's a contact to a potential person/group that I'm trying to angel in, I put them into a "dealflow" group, and I label key messages like decks, round information, and relevant articles/resources with tags so that they are easily queriable, same for Calendly links and emails (I mirror these tags on email and DocuSign, so I keep track of the entire lifecycle). Whenever a new opportunity comes around, I log it in my notes/spreadsheets and label it with the appropriate status. If it's a deal I haven't closed yet, I will set up reminders every 3 days to follow up and message the respective founder teams for status updates and update the status with their information. This is invaluable in not letting things fall through the cracks. Telegram bots allow you to set reminders to write to a certain chat. You can also create a folder called "follow up" and check it at the end of every day or other day to see if there are any actionable items for you to do. Communications overhead will get tough at some point, so also explore getting an EA if your calendar starts clashing and you're not able to meet some of your other obligations. Money spent on anything that makes administrative tasks less burdensome or saves you time is money well spent because it will allow you to scale up your important work. I recommend tracking all of your deals using a combination of notetaking applications (I use Notion and Obsidian) / spreadsheet program and a portfolio tracker (I personally use Rotki Premium; Rotki is a local free OSS portfolio tracker). Rotki has a great feature, and that is "manual balances." you can create asset/liability types and label them, e.g., "Clique seed round equity" (you can also add notes on a per asset basis, e.g., "Clique raised $ X amount at $ Y valuation with this vesting schedule..., relevant documents are on DocuSign, etc."). I keep track of the current value of my investments by updating them whenever portcos raise follow-up rounds or the price of the shares on secondary markets, and when they launch a token, I update the asset to just reflect the value denominated in the token (Rotki uses a variety of APIs so you should see their token there automatically or you can add a price feed API manually). This will track the real-time valuation and help you manage your assets effectively. You can also track exchange balances, bank accounts, and on-chain balances across many, many chains, etc. Definitely highly recommend it, and if you have a chance, give them a grant (Gitcoin, OP RPGF, or directly) since they are building all of this as a free OSS public good.

This information will help you immensely with accounting tax obligations when time goes around to selling and to just have peace of mind that you know that you are on top of things and that you're not letting any opportunities slip through the cracks. Founders are appreciative of proactive investors as it saves them time chasing people around and stressing about not having SAFEs/docs signed and money received, they have enough stress as it is. Whenever you receive a SAFE or similar document, review it carefully and make sure you understand the terms of your investment, like vesting schedules, valuations, token conversions through token side letters/warrants, or if it's just equity, etc etc. Modern LLMs and search engines like Perplexity are extremely helpful to get a basic education. If you want to be sure, contact a lawyer to help you review these documents.

I also wanted to shed some light on how I think about capital allocation. Since I'm very young and I purposefully live well below my means, I am able to save most of my monthly earnings and have also made a substantial amount of money from airdrops and from being in crypto over the span of 6 years (I started when I was 17 and about to enter my junior year of high school - I bought my first eth at around $200 in 2018). I used to mostly hold LST ETH and a few tokens I liked, but ever since I started angel investing, I have liquidated all of my liquid assets into stablecoins, started saving up my income every month, and put it all into angel investments over time. This is the sole reason that I have been able to do so many angel rounds where the ticket size goes anywhere between $10-75k, with around $20k being the average amount. If you are a builder like me or someone who doesn't do research or things that give you the edge in the liquid market, then angel investing might be the best source of asymmetric risk/reward opportunities aside from equity/tokens in the projects you work on (which I believe are the things builders should prioritize and supplementing it with angeling if you want to). Usually, the lowest ticket sizes projects take are $10k (or $5k if it's a pre-seed/seed in some cases), so bear this in mind as you start your angeling journey. Also, keep in mind angel investments are highly illiquid for several years and usually have very little liquidity on secondaries unless the projects are very hot or you make a substantial discount over the spot price if you have a long lockup.

Closing thoughts

I hope this article has helped you in some way, if you have any questions or suggestions, feel free to reach out to me on X or Telegram. Angel investing, to me, is an important way to build relationships with people I admire, enjoy spending time with, and would like to work with because I believe in their missions, values, and products, as well as in their ability to pull it off. It is a great way to share ownership over the technologies that we are building and to asymmetrically benefit when you know where the wind is blowing.

I will keep updating my portfolio companies on my website with a short blurb on why I like them and what they are doing.

Thank you, cat @pseudotheos, for helping review my article, me and mea familia will be forever grateful.

What's next for me

Phew, that was a lot to write. Now, we are getting to the good stuff, and that is where I want to head next. I feel like over the last 6 years, I’ve gone very deep into crypto, which has expanded my knowledge of our industry and has taught me a lot of things, but it has also narrowed my horizons. I’d like to learn a lot more about the world beyond just crypto because I want to build useful things for everyone, which is one of the main reasons I am working on Worldcoin. This August, I have planned to spend about 3 weeks in the Bay area in order to meet more people outside of my realm of specialty, I want to meet people into AI, biotech, physics, engineering, space technology, longevity, energy, hardware and lots of other domains in order to expand my world view. I want to spend more time around founders of a different nature and learn more about their fields of expertise. This doesn’t mean I want to quit crypto, not by a long shot, but I do want to build tools that are useful for everyone, just like longevity, AI, or hardware generally are used by everyone. In the immediate short term, I will close my current deal flow, and after I do so I will be very happy with my crypto portfolio, but I will be looking to angel invest outside of crypto as well once I have more liquidity to deploy.

I hope to catch up with you all at ETHCC in Brussels, SBC, Zuzalu, or Devcon in Thailand later this year, friends.

DC