Midnight is the Cardano-aligned, privacy-first sidechain engineered by IOHK / IOG. As a brand-new public chain, every line of code, every contract, every distribution event would be examined publicly — and would handle real value at scale.

We were brought in to design and run the quality, compliance and data programme that would let the team move fast without compromising the trust the chain’s name was being built on.

We built and maintained automated test suites covering protocol interactions, on-chain data flows, transaction lifecycles and user-facing surfaces. Pipelines were wired up in GitHub Actions as a first-class part of the development and delivery process — every change ran the relevant suites, and green or red was a clear, fast signal instead of a vibe.

• API and contract-level automation
• End-to-end transaction and lifecycle automation
• Regression suites that grew with the protocol
• GitHub Actions pipelines feeding the development & delivery flow
• Stable, reviewable test code — treated like product code

The Glacier Drop targeted holders across eight chains — Bitcoin, Ethereum, Cardano, Solana, BNB Chain, XRP Ledger, Avalanche and BAT — based on a snapshot of qualifying wallets, with allocations unlocking on a 360-day thawing schedule. We architected the data testing programme behind it: ingesting source records, running ETL transforms, reconciling allocations against eligibility rules and verifying the on-chain output.

• End-to-end ETL validation across source, staging and chain
• Allocation & eligibility reconciliation against rule sets
• Defect triage on the data layer with full audit trails
• Sign-off evidence for every batch released

The scale and shape of the data didn’t fit any off-the-shelf tool. So we wrote our own — a suite of custom Python utilities to ingest, transform, reconcile and spot-check enormous datasets, with reproducible output and clean reports for engineering, compliance and leadership.

• Bulk reconciliation across millions of records
• Automated diffing between source, staging and chain
• Reusable test fixtures & data generators
• Dashboards and CSV/JSON evidence outputs

We worked side-by-side with the UX/UI designers building the user-facing app — helping shape requirements before they reached engineering, instead of validating them after the fact. Two questions ran through every design review: is it accessible, and is it testable?

• Co-authored requirements for user-facing flows
• Accessibility built in — semantic structure, keyboard paths, screen-reader labels, contrast
• Testability built in — stable selectors, predictable states, well-defined empty / loading / error views
• Acceptance criteria that designers, devs and QA could all sign off on

The result: a user-facing experience that was inclusive by design and testable by default — instead of the usual last-minute scramble to retrofit either.