Protected Case Study
This portfolio piece is shared selectively. Enter the access code to continue.
Incorrect code — try again.
B2C surface shown here · B2B & B2E available on request
The Challenge
Interglobe Aviation (IndiGo Airlines) · Consultant Digital
Four journeys — booking, check-in, IFE, BluChip — grown independently, inconsistent, ungoverned. A design org fragmented across pods, agencies, and contractors, each with its own rituals.
The mandate: reset all priority journeys, build the ops model that sustains them, and do it at airline pace.
Approach
Discovery and delivery ran in lockstep: compressed BRD-to-wireframes, and hardened the DesignOps model that made speed sustainable. Both visible in the AI delivery section.
What I drove
Journeys
Problem: Booking, check-in, IFE, and BluChip had grown independently — inconsistent patterns, no shared review gate.
Action: Set a common interaction language, tightened funnel states with product and front-end leads, shipped in stages.
Result: Fewer support escalations, cleaner handoffs per sprint, a testable base for A/B experimentation.
DesignOps
Problem: No shared cadence, no utilization visibility — partners operating to different quality bars.
Action: Daily DSM syncs, mandatory peer reviews, utilization dashboard, quality checkpoints extended to all partners.
Result: 250% improvement in delivery turnaround. Rework dropped because reviews caught issues early.
Design System
Problem: Legal required distinct identities for flight vs. marketplace — no shared token layer, no developer handoff path.
Action: Built token-driven Figma DS in 4 weeks — then extended it as a CLI/NPM package so developers and AI agents could generate compliant UI in real time.
Result: Design shifted to DAT (Design Assessment Test) — validate output, measure the gap, report. Governance as a CI loop, not a gate.
Voice of Design
Problem: Post-delivery feedback was hallway comments and unthreaded chat — no structured signal.
Action: Introduced Design CSAT — product, business, and engineering score every major delivery.
Result: Evidence-based iteration; data became the basis for headcount and tooling investment.
Organization
Problem: Pods, agencies, TnM contracts — each with own rituals, file structures, and definition of done.
Action: Shared libraries, single review committee for major programs, structured onboarding for new partners.
Result: Partners ramped faster; quality stopped depending on which pod picked up the work.
Innovation
Problem: AI untested against IndiGo's real constraints — brand governance, regulatory copy, multilingual requirements.
Action: Focused pilots: layout exploration, research synthesis, prototype generation — each with success criteria and a human review gate.
Result: A working playbook for where AI accelerates vs. where it creates risk at regulated brand scale.
Featured · InFlight Entertainment
Designed for the in-flight context — content-rich, offline-aware, and gesture-native. One of four phygital surfaces redesigned under a single interaction language.
Marketplace Design System
A legal requirement — flight and marketplace must look distinct — became the forcing function to build IndiGo's first developer-facing design system. Built in Figma, then extended as a CLI/NPM package so developers and AI agents could generate compliant UI in real time. Design's job shifted from gating to assessing.
marketplace-hero.png to populate.
Marketplace screens available on request — drop marketplace-02.png and marketplace-03.png to populate component and token views.
Design Solution
Three layers of the full service blueprint. Download full UJM (PDF)
AI-First Delivery
Two recordings of real AI-assisted workflows in action — not slides. Not prototypes. Actual delivery.
How AI-accelerated component thinking and design-to-dev handoff compressed the Figma-to-production cycle inside the Catering Ops program — closing the design-engineering gap at delivery pace.
How AI-assisted synthesis moved from business requirements through logic, user flows, and IA to wireframes — compressing a process that used to take weeks into a fraction of the time.
Business Impact
"Governance wasn't bureaucracy — it was how we bought back 250% speed without lowering the quality bar."
DesignOps cut review cycles and eliminated rework. Quality caught early, not after handoff.
Four high-stakes flows redesigned and shipped at airline cadence — shared interaction language, dev-ready handoff.
Marketplace DS in 4 weeks — token-driven, Figma-native foundation for every future build from day one.
Pods, agencies, contractors unified under one model. Quality stopped depending on which pod picked up the work.
The 250% figure reflects internal turnaround metrics post-DesignOps; detailed baselines remain confidential.
AI-First Practice
Personal builds that sharpen the same muscles hiring teams care about — Figma-to-production discipline, metrics-grounded rationale, and assistant-shaped interfaces.
Tooling to ingest analytics reports and use AI-assisted workflows for revamp vs. greenfield decisions.
CLI experiments generating DS artifacts from Figma components — tightening the designer-developer handoff loop.
SwiftUI usability-test build aligned to Figma tokens — velocity and craft for airport-ready sessions.
Voice-reactive interfaces with speech-to-text — exploring patterns from conversational search and AI assistants.
Each build uses a distinct AI-assisted workflow developed over the past year. Details kept for in-person conversations — they spark the most interesting interview discussions.
Shared Patterns
The design problems are structurally the same — only the brand changes.
shreeransubhe.design@gmail.com · +91 98333 99080 · LinkedIn · shreeransubhe.design
