Intelligence Track

Design

25 June 2026

The Brief

The most consequential distribution move today isn't on any booking platform — it's a carrier redefining itself as a transit hub. Air India's Easy Connect lets a Varanasi passenger clear immigration at home and treat Delhi as an international transfer point, aimed squarely at the roughly 17 million Indians who route through Dubai, Doha or Singapore each year, folding multi-leg journeys into one airline's ticket. Read across the rest of the day, the same move repeats: a government super-app collapsing reserved, unreserved and platform rail ticketing into one login; a hotel-tech vendor arguing personalisation should let properties own willingness-to-pay before a comparison screen loads; a design tool absorbing code, motion and AI image editing so the handoff between design and engineering disappears. The common dynamic is integration eating intermediation — every incumbent here is pulling a layer that used to be someone else's business back inside its own walls, while a fourth carrier scaling profitably toward 40% international keeps adding the direct supply that makes aggregation less scarce. The question to sit with: when suppliers own the full journey and infrastructure providers own the booking surface, what does the aggregator do that neither can absorb — and is Cleartrip building that, or deferring it?

HeyDesigner · 25 Jun 2026

Six canvas additions were announced: code layers that turn any frame into editable React-backed code and clone GitHub repos onto the canvas, a native motion timeline exporting to CSS/React/MP4, AI-generated WebGPU shader fills and effects, prompt-built generative plugins, Weave AI image tools inside Design files, and an upgraded agent with reusable skills and connectors to Notion, GitHub and others — treating code as a design material equal to vectors.

Industry lens

Do product teams actually adopt code layers as a shared design-engineering surface, or keep routing interactive work to standalone coding agents — the behaviour that decides whether the canvas absorbs the workflow or gets bypassed?

Reading as

AI & Design

Your design system's newest author is an agent

A follow-up to last year's 'agent as design-system reader' argument documents that agents now write to design systems — placing real components on the Figma canvas via use_figma, proposing and committing token changes through Storybook and Figma MCP servers, generating documentation, and even authoring the SKILL.md and DESIGN.md files that instruct them — removing the human translator between system and implementation.

Once agents author components, tokens and docs at machine speed, the load-bearing failure shifts from generation quality to review: the sign-off that used to sit between a change and what ships becomes nobody's explicit job, so teams need a review and provenance model built for probabilistic authorship before they widen agent write access.

HeyDesigner·25 Jun 2026
Personalization: hospitality’s most powerful driver of revenue growth

An Amadeus executive, citing the vendor's Travel Dreams 2026 research, argues travellers pay more when relevance is delivered consistently from the search-and-compare stage through in-stay service, framing AI-plus-connected-data personalisation as a direct lever on conversion, spend and repeat booking rather than a loyalty nicety.

If hotels can lift willingness-to-pay by recognising a guest's purpose and budget before booking, the value-creating moment moves upstream to whoever holds the guest data and the pre-stay surface — and a claim made by a major hospitality-tech supplier doubles as a pitch to keep that data and surface hotel-side rather than with the intermediary.

Hospitality Net·25 Jun 2026