SSV Labs

Problem

BloxStaking ran a freemium, AWS-managed staking service. SSV Labs was migrating that entire user base to SSV Network, a decentralized validator infrastructure (DVT) network where the same validators would now be split across multiple independent operators instead of one managed instance, and users would pay operators directly instead of getting the service for free.

That's not a UI change. It's a security-critical, economically-different migration: users had to terminate their existing AWS setup, generate new key material, wait out a mandatory inactivity period to avoid getting slashed, understand a new token-based fee model, and register their validators on-chain, all without losing a single validator to a mistake mid-migration. 450+ operators, 2,500+ Ethereum validator nodes, $400M+ in staked assets. No margin for a bad instruction.

What I Built

I personally authored the official BloxStaking → SSV Network migration guide, the resource every one of those 450+ operators used to move their validators without getting slashed. In parallel, I built the integration governance framework that kept the partner-facing side of that migration from breaking anything downstream.

Both were the same underlying problem: translate a process with real financial and security consequences into something a non-expert could execute correctly, once, without a support ticket.

When a mistake costs someone their staked ETH, the guide isn't documentation on the side. It's the product.

Approach

The migration guide had to carry three kinds of weight at once. Security-critical sequencing: validators had to sit inactive for a minimum of 2 epochs (~13 minutes) before re-registering, verified against beaconcha.in, and getting this wrong and the validator gets slashed. New token economics: SSV Network's fee model has three distinct costs (operator fee, network fee, liquidation collateral), and users needed a formula (Total SSV = Operator Cost per Validator per Year × Number of Validators) to know what they'd actually pay, a freemium user's first encounter with a paid, market-priced service. Operational edge cases: a 500-validator-per-operator capacity ceiling meant some operators had to split their registration across multiple transactions, so the guide had to catch that before someone hit the wall mid-migration.

On the partner side, I ran the same discipline applied to internal squads: audited where escalations were actually coming from across 450+ operators, traced it to ambiguous integration contracts, and wrote governance covering functional requirement, system context, edge conditions, and acceptance criteria, applied across 8+ partner integrations. In parallel, I led the $400M+ asset migration across 4 engineering squads, scoping stability requirements first so live operators and downstream partners never felt disruption during the transition.

Also built Python + Cron automation for validator on/offboarding, replacing a manual, error-prone process across all 450+ operators, instrumented with pre/post KPI baselines before optimizing further.

PM Skills Applied

Security-critical technical writing for a non-technical, financially-exposed audience. One wrong instruction in the migration guide risks a user's staked ETH. Token economics translated into a formula a user could actually apply before committing funds. Cross-functional governance without direct authority over engineering squads. Compliance communication across B2B and B2C partner channels: cryptographic key management requirements, slashing condition constraints, and network security thresholds, translated into language both partners and engineers could act on.

Technical Depth

Wrote and tested the migration flow across Mainnet, Goerli, and Holesky before it reached real users. Documented the full keystore-to-keyshare pipeline (Wagyu Key Gen GUI and Staking Deposit CLI paths, ssv-keys splitting), the two-transaction on-chain registration flow (SSV spend approval, then validator registration), and the slashing-prevention verification step against beaconcha.in. Python + Cron automation scripts for validator on/offboarding at scale. Integration contracts covering functional requirement, system context, edge conditions, and acceptance criteria. KPI instrumentation built pre-launch, not bolted on after.

Lessons

---

The Story

Distributed systems fail quietly until they don't. At SSV Labs, the real risk wasn't a single bad integration. It was 450+ operators, each personally responsible for their own validators, all needing to execute a security-critical migration correctly on the first try. I didn't fix that with more process alone. I wrote the guide that made the safe path the only path anyone took, then proved the governance held: 50% fewer partner escalations, zero breaking changes across three major releases, and 2,500+ validator nodes moved across a $400M+ migration without a single slashing incident.