Case Study 02 / 05 · Accessibility · Pipeline Design · AI Collaboration

Fast A11y Pipeline

Built a pipeline to unblock the a11y backlog without pulling engineers off their roadmap. Won People's Choice.

February 2026 → Ongoing·First PR ready·Hackathon win

Problem

Accessibility gaps were blocking revenue. Engineers didn't have bandwidth to fix the backlog, and the longer it sat, the more it compounded. I needed a way to unblock the work without waiting on engineering capacity.

What I did

I built a pipeline and got designers and engineers on board. Design ships intent packs, Claude Code generates the fixes and opens PRs with before/after screenshots, and Engineering reviews and merges. I streamlined Claude Code into the workflow so the a11y backlog could move without pulling engineers off their roadmap.

How AI was used, with guardrails

Claude Code runs the scans, generates diffs, and opens PRs. All the labor. But I built the pipeline so AI is structurally blocked from making authority decisions: no auto-merge, no unsupervised fixes, no AI-decided severity levels.

Where human judgment mattered

I designed the authority boundary so AI never makes judgment calls. Design defines the intent (heading hierarchy, label copy, announcement language), Engineering gates every merge, and ambiguous cases always route to a human. The pipeline enforces this structurally, not by trusting AI to know its limits.

Relevant detail

Won People's Choice at a company-wide hackathon in February 2026. I negotiated a 1–2 day PR review cadence with engineering leads, and every PR ships with before/after screenshots. Separate ticket tracks per product surface.

1–2dPR review cadence
3Fix tiers
100%PRs with screenshots

Authority Distribution

Design
Intent
AI (Claude)
Labor
Engineering
Merge
Ambiguous
→ Human
Hackathon Win

People's Choice at company-wide hackathon, February 2026. Proved the pipeline model in 48 hours.

Pipeline Design

Three-role separation: Design ships intent packs → AI generates diffs → Engineering gates merges.

Tier Taxonomy

Tier 1: auto-fixable (alt text, aria-labels). Tier 2: needs design input. Tier 3: architectural, always human.

Production Cadence

1–2 day review loop negotiated with engineering leads. Before/after screenshots on every PR. Two product tracks running parallel.

A11yPipelineHackathonEnterprise

Get in touch

For detailed case studies, reach out.

hello@tinasingh.app