Vlad Velmeshev
4 min readJul 7, 2022


How the latest trends in crypto can help form an investment thesis.

Since the pandemic restrictions have been lifted worldwide, the crypto community flocked back to conferences for education and in a search for alpha. As a Partner at a crypto fund, I’ve also actively traveled in the past six months amidst ever-deteriorating market conditions. I found these events influential in shaping my sense of direction for blockchain technology and our investment thesis.

More recently, I’ve observed the following two key trends:

a) A lot of work is being done at the middleware layer. Think of it as a fabric that connects an underlying L1 infrastructure. It is, in effect, L2, but L2, which spans multiple chains (so I will call it L2, as existing scaling solutions will eventually span across multiple networks). This technology can only be powered by complex cryptography (zk and optimistic roll-ups) and advanced blockchain indexing protocols. Killer dapps of the future will sit at the layer above (L3) and communicate with L1s through middleware. These dapps can leverage the most optimal infrastructure solutions for a given task. E.g., if the dapp is related to loyalty points, NFT minting, or gamified experience — it might use Arweave for storage, identity management on Oasis, NFT minting on Solana, Liquidity pools on Etherium, etc. It will have to communicate with these protocols through cryptographic proofs to achieve speed and scalability. As it stands, Zk technology (Starkware, ZkSynch, Nil Foundation) is more promising in the long run, while optimistic roll-ups (Optimism, Arbitrum, Boba, Metis) are already operational. They can now be used for more straightforward use cases (with a limitation of optimistic assumptions and a transaction challenge period of up to 7 days for Optimism).

L3 to L1 communication requires advanced blockchain indexing protocols, which should save the blockchain information and information from dapps executed at L1 in the form of cryptographic proofs.

Finally, Celestia and Fuel (if proven to be operational) are going to play an essential part in this L1/L2/L3 paradigm, as Celestia is already optimized for computation and will have no execution at L1. Fuel is positioning itself as an execution layer or, in effect, the middleware mentioned above.

Technology will initially go through the stage where there will be at least five key players at each level because these teams and foundations are limited by the number of projects they can integrate into their protocols at any given time. Due to the network effects, this number will likely reduce in the long run. From an investment point of view, it makes more sense to spread the bets. Kosmos VC is well-positioned by having exposure across all top L1 projects and will continue accumulating positions at L2 — middleware layer.

It might take up to two-three years for middleware to become developed enough to support the dapps with millions of users, aka killer dapps.

b) A massive wave of consumer apps is coming into the space. The latest move-to-earn examples are the first glimpse of this tsunami. While, to date, play-to-earn and move-to-earn projects have been based on Ponzi tokenomics, real killer dapps will bring millions of users onchain without a compromise for sustainable token models. A generation of young, innovative, experienced traditional business owners is converting to the Web 3.0 paradigm. All of them are scrambling for a guide to the best infrastructure integration. These founders go from foundation to foundation, looking for tokenomics and game theory experts, skilled professionals who can help translate their business requirements into executable solutions powered by blockchain technology.

It is now common to call these projects Web 2.5.

I think that gamification of loyalty points, ticketing, and community rewards might become the first case for killer web 2.5 dapps in the near future.

Kosmos VC Investment thesis:

How do these two trends relate to our investment thesis?

I will digress here into a conversation we had with Ekram Ahmed — a pretty legendary guy in the closed circles who has helped position the brands for Celestia, Fuel, and Kleiner Perkins in the past. We discussed how to position Kosmos VC as a brand. The main task he gave us was to think about ONE thing we want to be best at and sacrifice everything else for it.

And the answer is: We want to be the best partner for integrating the killer dapps into the blockchain infrastructure. To bake this cake, we need to have the following ingredients:

a) L1 exposure and direct access to Foundations and developer relations teams — we’ve done pretty well here!

b) L2-middleware exposure and direct access to Foundations — this is a work in progress.

c) Experts in tokenomics (mathematic modeling and game theory) and cryptography — currently, we rely on third-party partners for these skills but are planning to bring them in-house.

That’s it! Sacrifice everything else for it. We will continue nourishing our developer community and hackathon relationships, marketing and PR contacts, exchange and launchpad listings, and market-making partners — but this is a complimentary service. Our core edge is cryptography and token economics design — the essential skills that are going to help experienced Web 2.0 founders transition into the promised land of Web 3.0