Generative AI is changing how design gets done: Flipkart's VP at Ahmedabad Design Week
At Ahmedabad Design Week, Gaurav Mathur, Vice President of Design at Flipkart, said the next big platform shift is here. We're moving out of the smartphone era and into the generative era - and product teams will feel it first.
His core point was simple: access beats novelty. Generative tools live in the browser, cost pennies to use, and compress time-to-first-draft to minutes. "To create an image using generative AI, it costs less than one rupee," he said. That drop in cost and friction is reworking creative workflows and production models across the board.
Teams without walls
Generative AI isn't "replacing designers." It's redistributing work. Mathur noted the old boundaries are fading: a product manager can design, a designer can ship code, an engineer can craft UI. Outputs won't always be perfect, but the direction is set.
He described the future team as "makers" - people with deep strength in one area and functional fluency across others. Less back-and-forth handover. More momentum.
What gets automated vs. what stays human
- Automated: layouts, icons, UI screens, GUI elements, quick prototypes. The grunt work moves to machines.
- Human-critical: selecting the best option, spotting flaws, seeing patterns, and aligning choices with behavior and culture. That's judgment, not generation.
For product leaders, the operating question shifts from "Who can build this?" to "Who can choose, refine, and ship the right thing at speed?"
Skills product teams need next
- Critical thinking: evaluate AI outputs, pressure-test edge cases, make trade-offs explicit.
- Cultural sensitivity and empathy: design choices land differently across contexts; AI won't intuit that.
- Behavior insight: map mental models, friction points, and motivation - then tune prompts and flows accordingly.
- Business literacy: Mathur was blunt: designers and PMs must understand acquisition, retention, unit economics, and go-to-market. Otherwise, great prototypes stall.
Practical moves for product leaders
- Redesign roles: hire and grow "T-shaped" makers. Update job ladders to reward cross-functional fluency.
- Change the workflow: make generation cheap and frequent; formalize gates for human review and decision.
- Add AI QA: build prompts, critique rubrics, and checklists for accessibility, cultural fit, and brand voice.
- Instrument the loop: track option-to-selection ratio, time-to-iteration, and user impact from AI-assisted work.
- Create a safe stack: approved tools, data boundaries, and a lightweight governance doc everyone can follow.
- Upskill continuously: short sprints on prompting, interface heuristics, and business basics for the entire pod.
How to pilot inside your org (4-week plan)
- Week 1: pick one flow (onboarding, search, or checkout). Define quality criteria. Set up a shared prompt library.
- Week 2: generate 20-50 variants. Run quick user tests. Keep decision logs on why options were kept or cut.
- Week 3: converge on 2-3 candidates. Ship an A/B or feature flag to a small cohort.
- Week 4: review metrics and support tickets. Document what the AI got wrong and how the team caught it. Roll learning into the next cycle.
Why this shift matters now
Previous platform changes - smartphones and cloud - rewired product work. Generative AI is doing the same, with faster feedback and far lower creation costs. Access is widespread, so advantage goes to teams who can decide, not just produce.
Ahmedabad Design Week is being hosted at Karnavati University from January 30 to February 1, bringing together designers, industry leaders, and academics to discuss AI's impact across sectors.
If you want a broader view of how generative AI is progressing, the latest AI Index offers useful context: Stanford AI Index.
For structured upskilling by role (PM, designer, engineer), see curated learning paths here: Complete AI Training - Courses by Job.
Your membership also unlocks:
