Soda Labs raises $3 million from one backer to hide bank balances on public blockchains

NextBlock opened its first alternative investment fund in Luxembourg in July 2025, starting out with $40 million committed ^[9]. If Soda's check came from that pool, it is 7.5% of it ^[21], a large position for one investor to hold alone at seed. Pieter van Poecke, NextBlock's founder and general partner, said "Soda already had a working product and paying customers" ^[8]. He also credited the founding team's technical IP and "strong commercial instincts" ^[19].

Cryptopolitan describes the raise as paying for Bubble's expansion from COTI onto those four chains and eventually Solana ^[2], yet the layer is already running there ^[7]. Most of the money, then, pays for selling. The stated use is commercial adoption over the next 12 to 18 months ^[4], along with the move from Soda's older gcEVM to Bubble ^[6]. Bubble is a chain-agnostic coprocessor meant to process private workloads from several chains without exposing the data to outsiders or to Soda itself ^[6].

Avishay Yanai, Soda's co-founder and chief executive, said: "This round lets us take it from pilots to production." ^[5] Soda already runs in production on COTI, for about 19 months as of the October 2026 close ^[24]. The pilots Yanai means are presumably the bank and payment-firm deployments Soda wants on public chains, where the aim is to keep balances, amounts and counterparties private ^[3].

Bubble splits encrypted data across independent nodes so that no single operator can read it, and it runs on ordinary cloud CPUs with standard cryptography such as AES and SHA256 ^[10]. Amounts, token types, senders and receivers stay off the public record, and an authorized party can still decrypt them ^[11].

The performance record is thin. At 500 confidential transactions a second, the 15,000 Arbitrum transfers behind the benchmark on Soda's website amount to about 30 seconds of sustained load ^[23]. Soda says it has since improved full-lifecycle performance five- to tenfold and will publish the new numbers in the coming weeks ^[12].

The good outcome has Soda naming a bank or payment firm that moves real balances through Bubble on Ethereum or Base before the money runs out ^{[7] [4]}. The middle one has the pilots staying pilots while Soda raises again on its COTI record ^[15]. In the bad one, buyers go to rivals such as EthSystems, which former Ethereum Foundation researchers launched in July 2026 to pitch privacy infrastructure to banks ^[18]. 21Shares called privacy "the missing layer between blockchains and institutions" in September 2026 ^[16].

I think the middle outcome is the likeliest on this evidence. The announcement does not name the paying customers van Poecke cited ^[8], and the public benchmark covers half a minute of traffic ^[23]. The case against that view is the round's size: $3 million ^[1] is a reasonable sum for a team with a live product to spend converting customers it already bills, if van Poecke's description is right ^[8]. A named bank or payment firm settling regulated money on a public chain through Bubble within 18 months ^[4] would show the bet was good and my view wrong.